covid-19-transfer-learning-resnet152v2.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# COVID 19 CLASSIFICATION\n\nDone by: Sebastián Sarasti","metadata":{}},{"cell_type":"markdown","source":"The COVID-19 disease provoked a global pandemic. It was an important worry around the world.\n\nIt was the reason why many people around the world passed away. Many stratigies were purposed to address COVID-19 and reduce the impact.\n\nOne way to diagnosticate people is by X-ray scans like this:\n\n![dataset-cover (1).png](attachment:bc578edc-3861-4a12-969d-9be15f8a3fbf.png)\n\nA possible option is to create a Convolutional Neural Network and try to predict the COVID-19 disease based only in the X-ray scan. ","metadata":{},"attachments":{"bc578edc-3861-4a12-969d-9be15f8a3fbf.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"This data set has x-ray scans labeled into three categories: covid, normal, and viral pneumonia.\n\nThe goal is generate an algorithm which is able to classify between each category","metadata":{}},{"cell_type":"markdown","source":"*Import basic libraries*","metadata":{}},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-09-16T13:23:08.885315Z","iopub.execute_input":"2022-09-16T13:23:08.885714Z","iopub.status.idle":"2022-09-16T13:23:08.920657Z","shell.execute_reply.started":"2022-09-16T13:23:08.885626Z","shell.execute_reply":"2022-09-16T13:23:08.919383Z"},"trusted":true},"execution_count":1,"outputs":[]},{"cell_type":"markdown","source":"*Import libraries to load the data*","metadata":{}},{"cell_type":"code","source":"\nimport tensorflow as tf\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:23:08.925434Z","iopub.execute_input":"2022-09-16T13:23:08.925776Z","iopub.status.idle":"2022-09-16T13:23:17.275345Z","shell.execute_reply.started":"2022-09-16T13:23:08.925734Z","shell.execute_reply":"2022-09-16T13:23:17.273942Z"},"trusted":true},"execution_count":2,"outputs":[]},{"cell_type":"code","source":"train_gen = ImageDataGenerator(rescale=1/255,validation_split=0.2)\ntest_gen = ImageDataGenerator(rescale=1/255)\ntrain_data = train_gen.flow_from_directory(directory='../input/covid19-image-dataset/Covid19-dataset/train', \n                                           target_size=(300,300), subset='training')\nval_data = train_gen.flow_from_directory(directory='../input/covid19-image-dataset/Covid19-dataset/train', \n                                           target_size=(300,300),subset='validation')\ntest_data = test_gen.flow_from_directory(directory='../input/covid19-image-dataset/Covid19-dataset/test', \n                                         target_size=(300,300))","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:23:17.277465Z","iopub.execute_input":"2022-09-16T13:23:17.278446Z","iopub.status.idle":"2022-09-16T13:23:17.613118Z","shell.execute_reply.started":"2022-09-16T13:23:17.278408Z","shell.execute_reply":"2022-09-16T13:23:17.611780Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stdout","text":"Found 201 images belonging to 3 classes.\nFound 50 images belonging to 3 classes.\nFound 66 images belonging to 3 classes.\n","output_type":"stream"}]},{"cell_type":"markdown","source":"*Verify the data loaded*","metadata":{}},{"cell_type":"code","source":"train_data.class_indices.keys()","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:23:17.616726Z","iopub.execute_input":"2022-09-16T13:23:17.617222Z","iopub.status.idle":"2022-09-16T13:23:17.630493Z","shell.execute_reply.started":"2022-09-16T13:23:17.617180Z","shell.execute_reply":"2022-09-16T13:23:17.628475Z"},"trusted":true},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/plain":"dict_keys(['Covid', 'Normal', 'Viral Pneumonia'])"},"metadata":{}}]},{"cell_type":"markdown","source":"### Create the CNN architecture","metadata":{}},{"cell_type":"markdown","source":"*Import the needed modules*","metadata":{}},{"cell_type":"code","source":"from tensorflow.keras.models import Sequential\nfrom tensorflow.keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Dropout, Dense, Flatten\nfrom keras.regularizers import l1, l2","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:23:17.632165Z","iopub.execute_input":"2022-09-16T13:23:17.633801Z","iopub.status.idle":"2022-09-16T13:23:17.642490Z","shell.execute_reply.started":"2022-09-16T13:23:17.633739Z","shell.execute_reply":"2022-09-16T13:23:17.640930Z"},"trusted":true},"execution_count":5,"outputs":[]},{"cell_type":"markdown","source":"*Define the CNN structure*","metadata":{}},{"cell_type":"code","source":"a=0.07\nmodel = Sequential()\nmodel.add(Conv2D(filters=256,kernel_size=(3,3),padding='same',strides=(1,1), activation='relu',input_shape=(300,300,3), \n                 kernel_regularizer=l2(a), bias_regularizer=l2(a)))\nmodel.add(MaxPooling2D(pool_size=(2,2)))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.3))\n\nmodel.add(Conv2D(filters=128,kernel_size=(3,3),padding='same',strides=(1,1), activation='relu',kernel_regularizer=l2(a), bias_regularizer=l2(a)))\nmodel.add(MaxPooling2D(pool_size=(2,2)))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.3))\n\nmodel.add(Conv2D(filters=128,kernel_size=(3,3),padding='same',strides=(1,1), activation='relu',kernel_regularizer=l2(a), bias_regularizer=l2(a)))\nmodel.add(MaxPooling2D(pool_size=(2,2)))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.3))\n\nmodel.add(Conv2D(filters=128,kernel_size=(3,3),padding='same',strides=(1,1), activation='relu',\n                 kernel_regularizer=l2(a), bias_regularizer=l2(a)))\nmodel.add(MaxPooling2D(pool_size=(2,2)))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.3))\n\nmodel.add(Conv2D(filters=128,kernel_size=(3,3),padding='same',strides=(1,1), activation='relu', \n                 kernel_regularizer=l2(a), bias_regularizer=l2(a)))\nmodel.add(MaxPooling2D(pool_size=(2,2)))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.3))\n\nmodel.add(Flatten())\nmodel.add(Dropout(0.3))\n\nmodel.add(Dense(10,activation='relu', kernel_regularizer=l2(a), bias_regularizer=l2(a)))\nmodel.add(Dropout(0.3))\n\nmodel.add(Dense(3,activation = 'softmax'))","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:23:17.644413Z","iopub.execute_input":"2022-09-16T13:23:17.645870Z","iopub.status.idle":"2022-09-16T13:23:21.706520Z","shell.execute_reply.started":"2022-09-16T13:23:17.645804Z","shell.execute_reply":"2022-09-16T13:23:21.705175Z"},"trusted":true},"execution_count":6,"outputs":[{"name":"stderr","text":"2022-09-16 13:23:17.780412: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:17.905785: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:17.909953: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:17.914620: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\nTo enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n2022-09-16 13:23:17.914986: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:17.919172: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:17.923493: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:20.752749: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:20.756972: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:20.761091: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-09-16 13:23:20.765074: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 15403 MB memory:  -> device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0\n","output_type":"stream"}]},{"cell_type":"markdown","source":"* See all the architecture and parameters of the CNN*","metadata":{}},{"cell_type":"code","source":"model.summary()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-09-16T13:23:21.708503Z","iopub.execute_input":"2022-09-16T13:23:21.709026Z","iopub.status.idle":"2022-09-16T13:23:21.721376Z","shell.execute_reply.started":"2022-09-16T13:23:21.708993Z","shell.execute_reply":"2022-09-16T13:23:21.719789Z"},"trusted":true},"execution_count":7,"outputs":[{"name":"stdout","text":"Model: \"sequential\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nconv2d (Conv2D)              (None, 300, 300, 256)     7168      \n_________________________________________________________________\nmax_pooling2d (MaxPooling2D) (None, 150, 150, 256)     0         \n_________________________________________________________________\nbatch_normalization (BatchNo (None, 150, 150, 256)     1024      \n_________________________________________________________________\ndropout (Dropout)            (None, 150, 150, 256)     0         \n_________________________________________________________________\nconv2d_1 (Conv2D)            (None, 150, 150, 128)     295040    \n_________________________________________________________________\nmax_pooling2d_1 (MaxPooling2 (None, 75, 75, 128)       0         \n_________________________________________________________________\nbatch_normalization_1 (Batch (None, 75, 75, 128)       512       \n_________________________________________________________________\ndropout_1 (Dropout)          (None, 75, 75, 128)       0         \n_________________________________________________________________\nconv2d_2 (Conv2D)            (None, 75, 75, 128)       147584    \n_________________________________________________________________\nmax_pooling2d_2 (MaxPooling2 (None, 37, 37, 128)       0         \n_________________________________________________________________\nbatch_normalization_2 (Batch (None, 37, 37, 128)       512       \n_________________________________________________________________\ndropout_2 (Dropout)          (None, 37, 37, 128)       0         \n_________________________________________________________________\nconv2d_3 (Conv2D)            (None, 37, 37, 128)       147584    \n_________________________________________________________________\nmax_pooling2d_3 (MaxPooling2 (None, 18, 18, 128)       0         \n_________________________________________________________________\nbatch_normalization_3 (Batch (None, 18, 18, 128)       512       \n_________________________________________________________________\ndropout_3 (Dropout)          (None, 18, 18, 128)       0         \n_________________________________________________________________\nconv2d_4 (Conv2D)            (None, 18, 18, 128)       147584    \n_________________________________________________________________\nmax_pooling2d_4 (MaxPooling2 (None, 9, 9, 128)         0         \n_________________________________________________________________\nbatch_normalization_4 (Batch (None, 9, 9, 128)         512       \n_________________________________________________________________\ndropout_4 (Dropout)          (None, 9, 9, 128)         0         \n_________________________________________________________________\nflatten (Flatten)            (None, 10368)             0         \n_________________________________________________________________\ndropout_5 (Dropout)          (None, 10368)             0         \n_________________________________________________________________\ndense (Dense)                (None, 10)                103690    \n_________________________________________________________________\ndropout_6 (Dropout)          (None, 10)                0         \n_________________________________________________________________\ndense_1 (Dense)              (None, 3)                 33        \n=================================================================\nTotal params: 851,755\nTrainable params: 850,219\nNon-trainable params: 1,536\n_________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"markdown","source":"*Define a callback to save the best model according to the accuracy*","metadata":{}},{"cell_type":"code","source":"from tensorflow.keras.callbacks import ModelCheckpoint\ncheck = ModelCheckpoint('covid_19.hdf5',monitor='val_accuracy',save_best_only=True)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:23:21.723362Z","iopub.execute_input":"2022-09-16T13:23:21.724200Z","iopub.status.idle":"2022-09-16T13:23:21.732449Z","shell.execute_reply.started":"2022-09-16T13:23:21.724139Z","shell.execute_reply":"2022-09-16T13:23:21.730801Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"markdown","source":"*Compile and train the CNN model*","metadata":{}},{"cell_type":"code","source":"model.compile(optimizer='adam',loss='categorical_crossentropy',metrics='accuracy')\nmodel.fit(train_data,epochs=5,validation_data=val_data,callbacks=[check])","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-09-16T13:23:21.739015Z","iopub.execute_input":"2022-09-16T13:23:21.739946Z","iopub.status.idle":"2022-09-16T13:24:51.749401Z","shell.execute_reply.started":"2022-09-16T13:23:21.739881Z","shell.execute_reply":"2022-09-16T13:24:51.747977Z"},"trusted":true},"execution_count":9,"outputs":[{"name":"stderr","text":"2022-09-16 13:23:24.232783: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n","output_type":"stream"},{"name":"stdout","text":"Epoch 1/5\n","output_type":"stream"},{"name":"stderr","text":"2022-09-16 13:23:29.652199: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005\n","output_type":"stream"},{"name":"stdout","text":"7/7 [==============================] - 33s 3s/step - loss: 40.5146 - accuracy: 0.5323 - val_loss: 38.3862 - val_accuracy: 0.4200\nEpoch 2/5\n7/7 [==============================] - 14s 2s/step - loss: 36.2881 - accuracy: 0.7363 - val_loss: 34.1112 - val_accuracy: 0.4400\nEpoch 3/5\n7/7 [==============================] - 14s 2s/step - loss: 31.8253 - accuracy: 0.8657 - val_loss: 30.0777 - val_accuracy: 0.4400\nEpoch 4/5\n7/7 [==============================] - 13s 2s/step - loss: 27.6331 - accuracy: 0.8458 - val_loss: 26.9424 - val_accuracy: 0.4400\nEpoch 5/5\n7/7 [==============================] - 14s 2s/step - loss: 23.6927 - accuracy: 0.8458 - val_loss: 24.1411 - val_accuracy: 0.4400\n","output_type":"stream"},{"execution_count":9,"output_type":"execute_result","data":{"text/plain":"<keras.callbacks.History at 0x7f1e715c0ad0>"},"metadata":{}}]},{"cell_type":"markdown","source":"*load the best model to see the capabilities to predict*","metadata":{}},{"cell_type":"code","source":"from tensorflow.keras.models import load_model\nfinal_model = load_model('covid_19.hdf5')","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:24:51.752190Z","iopub.execute_input":"2022-09-16T13:24:51.753222Z","iopub.status.idle":"2022-09-16T13:24:52.086165Z","shell.execute_reply.started":"2022-09-16T13:24:51.753144Z","shell.execute_reply":"2022-09-16T13:24:52.084840Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"markdown","source":"*See the results*","metadata":{}},{"cell_type":"code","source":"final_model.evaluate(test_data)","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-09-16T13:24:52.088172Z","iopub.execute_input":"2022-09-16T13:24:52.088688Z","iopub.status.idle":"2022-09-16T13:24:56.470604Z","shell.execute_reply.started":"2022-09-16T13:24:52.088644Z","shell.execute_reply":"2022-09-16T13:24:56.469111Z"},"trusted":true},"execution_count":11,"outputs":[{"name":"stdout","text":"3/3 [==============================] - 3s 739ms/step - loss: 34.1512 - accuracy: 0.3939\n","output_type":"stream"},{"execution_count":11,"output_type":"execute_result","data":{"text/plain":"[34.15118408203125, 0.39393940567970276]"},"metadata":{}}]},{"cell_type":"markdown","source":"## Partial Conclusion:\n\nEven though the CNN is big, it is not able to generalize and get good results with the test data.\n\nA possible reason might be that the CNN is overfitting in the training data set.\n\nA solution is to create a simple CNN and try to see if the CNN is able to generalize and get good results with the test data.","metadata":{}},{"cell_type":"markdown","source":"*New simple model*","metadata":{}},{"cell_type":"code","source":"model1 = Sequential()\n\nmodel1.add(Conv2D(filters=64,kernel_size=(3,3),padding='same',strides=(1,1), activation='relu',input_shape=(300,300,3)))\nmodel1.add(MaxPooling2D(pool_size=(4,4)))\nmodel1.add(BatchNormalization())\n\n\nmodel1.add(Flatten())\n\n\nmodel1.add(Dense(3,activation='softmax'))\n","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:24:56.472170Z","iopub.execute_input":"2022-09-16T13:24:56.473011Z","iopub.status.idle":"2022-09-16T13:24:56.526732Z","shell.execute_reply.started":"2022-09-16T13:24:56.472960Z","shell.execute_reply":"2022-09-16T13:24:56.525556Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"markdown","source":"*See the model architecture*","metadata":{}},{"cell_type":"code","source":"model1.summary()","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:24:56.528646Z","iopub.execute_input":"2022-09-16T13:24:56.529072Z","iopub.status.idle":"2022-09-16T13:24:56.538601Z","shell.execute_reply.started":"2022-09-16T13:24:56.529031Z","shell.execute_reply":"2022-09-16T13:24:56.537097Z"},"trusted":true},"execution_count":13,"outputs":[{"name":"stdout","text":"Model: \"sequential_1\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nconv2d_5 (Conv2D)            (None, 300, 300, 64)      1792      \n_________________________________________________________________\nmax_pooling2d_5 (MaxPooling2 (None, 75, 75, 64)        0         \n_________________________________________________________________\nbatch_normalization_5 (Batch (None, 75, 75, 64)        256       \n_________________________________________________________________\nflatten_1 (Flatten)          (None, 360000)            0         \n_________________________________________________________________\ndense_2 (Dense)              (None, 3)                 1080003   \n=================================================================\nTotal params: 1,082,051\nTrainable params: 1,081,923\nNon-trainable params: 128\n_________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"model1.compile(optimizer='adam',loss='categorical_crossentropy',metrics='accuracy')","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:24:56.540487Z","iopub.execute_input":"2022-09-16T13:24:56.544266Z","iopub.status.idle":"2022-09-16T13:24:56.557211Z","shell.execute_reply.started":"2022-09-16T13:24:56.544234Z","shell.execute_reply":"2022-09-16T13:24:56.555650Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"markdown","source":"*Train the model*","metadata":{}},{"cell_type":"code","source":"model1.fit(train_data,epochs=15,validation_data=val_data,callbacks=[check])","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:24:56.560061Z","iopub.execute_input":"2022-09-16T13:24:56.561204Z","iopub.status.idle":"2022-09-16T13:28:28.898575Z","shell.execute_reply.started":"2022-09-16T13:24:56.561161Z","shell.execute_reply":"2022-09-16T13:28:28.896930Z"},"trusted":true},"execution_count":15,"outputs":[{"name":"stdout","text":"Epoch 1/15\n7/7 [==============================] - 15s 2s/step - loss: 8.7381 - accuracy: 0.7463 - val_loss: 1.4420 - val_accuracy: 0.7400\nEpoch 2/15\n7/7 [==============================] - 13s 2s/step - loss: 1.6476 - accuracy: 0.9353 - val_loss: 2.1251 - val_accuracy: 0.6200\nEpoch 3/15\n7/7 [==============================] - 12s 2s/step - loss: 0.4108 - accuracy: 0.9652 - val_loss: 1.9146 - val_accuracy: 0.6800\nEpoch 4/15\n7/7 [==============================] - 13s 2s/step - loss: 0.0015 - accuracy: 1.0000 - val_loss: 2.0627 - val_accuracy: 0.6600\nEpoch 5/15\n7/7 [==============================] - 13s 2s/step - loss: 0.2264 - accuracy: 0.9851 - val_loss: 1.6465 - val_accuracy: 0.7200\nEpoch 6/15\n7/7 [==============================] - 12s 2s/step - loss: 4.8632e-08 - accuracy: 1.0000 - val_loss: 1.0829 - val_accuracy: 0.7800\nEpoch 7/15\n7/7 [==============================] - 13s 2s/step - loss: 0.0552 - accuracy: 0.9900 - val_loss: 1.2591 - val_accuracy: 0.8000\nEpoch 8/15\n7/7 [==============================] - 13s 2s/step - loss: 7.1170e-09 - accuracy: 1.0000 - val_loss: 1.5621 - val_accuracy: 0.8000\nEpoch 9/15\n7/7 [==============================] - 13s 2s/step - loss: 1.3935e-05 - accuracy: 1.0000 - val_loss: 1.6981 - val_accuracy: 0.8000\nEpoch 10/15\n7/7 [==============================] - 13s 2s/step - loss: 0.0014 - accuracy: 1.0000 - val_loss: 1.6389 - val_accuracy: 0.8000\nEpoch 11/15\n7/7 [==============================] - 13s 2s/step - loss: 1.0320e-07 - accuracy: 1.0000 - val_loss: 1.5007 - val_accuracy: 0.8200\nEpoch 12/15\n7/7 [==============================] - 13s 2s/step - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.4030 - val_accuracy: 0.8000\nEpoch 13/15\n7/7 [==============================] - 12s 2s/step - loss: 1.7080e-07 - accuracy: 1.0000 - val_loss: 1.3307 - val_accuracy: 0.8000\nEpoch 14/15\n7/7 [==============================] - 13s 2s/step - loss: 1.3166e-07 - accuracy: 1.0000 - val_loss: 1.2718 - val_accuracy: 0.8000\nEpoch 15/15\n7/7 [==============================] - 13s 2s/step - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.2225 - val_accuracy: 0.8000\n","output_type":"stream"},{"execution_count":15,"output_type":"execute_result","data":{"text/plain":"<keras.callbacks.History at 0x7f1e701cc210>"},"metadata":{}}]},{"cell_type":"markdown","source":"*Verify if the model was able to learn well*","metadata":{}},{"cell_type":"code","source":"model1.evaluate(test_data)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:28.900798Z","iopub.execute_input":"2022-09-16T13:28:28.901737Z","iopub.status.idle":"2022-09-16T13:28:31.691169Z","shell.execute_reply.started":"2022-09-16T13:28:28.901684Z","shell.execute_reply":"2022-09-16T13:28:31.689580Z"},"trusted":true},"execution_count":16,"outputs":[{"name":"stdout","text":"3/3 [==============================] - 2s 552ms/step - loss: 1.7628 - accuracy: 0.7121\n","output_type":"stream"},{"execution_count":16,"output_type":"execute_result","data":{"text/plain":"[1.762752890586853, 0.7121211886405945]"},"metadata":{}}]},{"cell_type":"markdown","source":"## Partial Conclusion:\n\nThe model again was not able to generalize the learning, so based on the current amount of data. The source of this error might be that data are not as large to extract enough information.\n\nIn the cases were is not available enough data, a solution is to use transfer learning","metadata":{}},{"cell_type":"markdown","source":"*Import the model from Keras*","metadata":{}},{"cell_type":"code","source":"from tensorflow.keras.applications import ResNet152V2","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:31.692855Z","iopub.execute_input":"2022-09-16T13:28:31.693570Z","iopub.status.idle":"2022-09-16T13:28:31.703574Z","shell.execute_reply.started":"2022-09-16T13:28:31.693518Z","shell.execute_reply":"2022-09-16T13:28:31.700873Z"},"trusted":true},"execution_count":17,"outputs":[]},{"cell_type":"markdown","source":"*Resize the data to fit into the ResNetV2 architecture*","metadata":{}},{"cell_type":"code","source":"train_gen_resize = ImageDataGenerator(rescale=1/255,validation_split=0.2)\ntest_gen_resize = ImageDataGenerator(rescale=1/255)\ntrain_data_resize = train_gen_resize.flow_from_directory(directory='../input/covid19-image-dataset/Covid19-dataset/train', \n                                           target_size=(224, 224), subset='training', color_mode = 'rgb', batch_size=32, shuffle=True)\nval_data_resize = train_gen_resize.flow_from_directory(directory='../input/covid19-image-dataset/Covid19-dataset/train', \n                                           target_size=(224, 224),subset='validation', color_mode = 'rgb', batch_size=32, shuffle=True)\ntest_data_resize = test_gen_resize.flow_from_directory(directory='../input/covid19-image-dataset/Covid19-dataset/test', \n                                         target_size=(224, 224), color_mode = 'rgb', batch_size=32, shuffle=True)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:31.707048Z","iopub.execute_input":"2022-09-16T13:28:31.708392Z","iopub.status.idle":"2022-09-16T13:28:32.037400Z","shell.execute_reply.started":"2022-09-16T13:28:31.708354Z","shell.execute_reply":"2022-09-16T13:28:32.036043Z"},"trusted":true},"execution_count":18,"outputs":[{"name":"stdout","text":"Found 201 images belonging to 3 classes.\nFound 50 images belonging to 3 classes.\nFound 66 images belonging to 3 classes.\n","output_type":"stream"}]},{"cell_type":"markdown","source":"*Avoid a retraining in the ResNet model*","metadata":{}},{"cell_type":"code","source":"pre_trained_model = ResNet152V2(include_top = False,input_shape =(224,224,3))\nfor layer in pre_trained_model.layers:\n    layer.trainable = False","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:32.039122Z","iopub.execute_input":"2022-09-16T13:28:32.039538Z","iopub.status.idle":"2022-09-16T13:28:41.882429Z","shell.execute_reply.started":"2022-09-16T13:28:32.039497Z","shell.execute_reply":"2022-09-16T13:28:41.881081Z"},"trusted":true},"execution_count":19,"outputs":[{"name":"stdout","text":"Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet152v2_weights_tf_dim_ordering_tf_kernels_notop.h5\n234553344/234545216 [==============================] - 4s 0us/step\n234561536/234545216 [==============================] - 4s 0us/step\n","output_type":"stream"}]},{"cell_type":"markdown","source":"*See the model architecture*","metadata":{}},{"cell_type":"code","source":"pre_trained_model.summary()","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:41.884188Z","iopub.execute_input":"2022-09-16T13:28:41.884649Z","iopub.status.idle":"2022-09-16T13:28:41.992039Z","shell.execute_reply.started":"2022-09-16T13:28:41.884590Z","shell.execute_reply":"2022-09-16T13:28:41.990842Z"},"trusted":true},"execution_count":20,"outputs":[{"name":"stdout","text":"Model: \"resnet152v2\"\n__________________________________________________________________________________________________\nLayer (type)                    Output Shape         Param #     Connected to                     \n==================================================================================================\ninput_1 (InputLayer)            [(None, 224, 224, 3) 0                                            \n__________________________________________________________________________________________________\nconv1_pad (ZeroPadding2D)       (None, 230, 230, 3)  0           input_1[0][0]                    \n__________________________________________________________________________________________________\nconv1_conv (Conv2D)             (None, 112, 112, 64) 9472        conv1_pad[0][0]                  \n__________________________________________________________________________________________________\npool1_pad (ZeroPadding2D)       (None, 114, 114, 64) 0           conv1_conv[0][0]                 \n__________________________________________________________________________________________________\npool1_pool (MaxPooling2D)       (None, 56, 56, 64)   0           pool1_pad[0][0]                  \n__________________________________________________________________________________________________\nconv2_block1_preact_bn (BatchNo (None, 56, 56, 64)   256         pool1_pool[0][0]                 \n__________________________________________________________________________________________________\nconv2_block1_preact_relu (Activ (None, 56, 56, 64)   0           conv2_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv2_block1_1_conv (Conv2D)    (None, 56, 56, 64)   4096        conv2_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block1_1_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_1_relu (Activation (None, 56, 56, 64)   0           conv2_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block1_2_pad (ZeroPadding (None, 58, 58, 64)   0           conv2_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_2_conv (Conv2D)    (None, 56, 56, 64)   36864       conv2_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv2_block1_2_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_2_relu (Activation (None, 56, 56, 64)   0           conv2_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block1_0_conv (Conv2D)    (None, 56, 56, 256)  16640       conv2_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block1_3_conv (Conv2D)    (None, 56, 56, 256)  16640       conv2_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_out (Add)          (None, 56, 56, 256)  0           conv2_block1_0_conv[0][0]        \n                                                                 conv2_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_preact_bn (BatchNo (None, 56, 56, 256)  1024        conv2_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv2_block2_preact_relu (Activ (None, 56, 56, 256)  0           conv2_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv2_block2_1_conv (Conv2D)    (None, 56, 56, 64)   16384       conv2_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block2_1_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_1_relu (Activation (None, 56, 56, 64)   0           conv2_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block2_2_pad (ZeroPadding (None, 58, 58, 64)   0           conv2_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_2_conv (Conv2D)    (None, 56, 56, 64)   36864       conv2_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv2_block2_2_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_2_relu (Activation (None, 56, 56, 64)   0           conv2_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block2_3_conv (Conv2D)    (None, 56, 56, 256)  16640       conv2_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_out (Add)          (None, 56, 56, 256)  0           conv2_block1_out[0][0]           \n                                                                 conv2_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_preact_bn (BatchNo (None, 56, 56, 256)  1024        conv2_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv2_block3_preact_relu (Activ (None, 56, 56, 256)  0           conv2_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv2_block3_1_conv (Conv2D)    (None, 56, 56, 64)   16384       conv2_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block3_1_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_1_relu (Activation (None, 56, 56, 64)   0           conv2_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block3_2_pad (ZeroPadding (None, 58, 58, 64)   0           conv2_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_2_conv (Conv2D)    (None, 28, 28, 64)   36864       conv2_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv2_block3_2_bn (BatchNormali (None, 28, 28, 64)   256         conv2_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_2_relu (Activation (None, 28, 28, 64)   0           conv2_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nmax_pooling2d_6 (MaxPooling2D)  (None, 28, 28, 256)  0           conv2_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv2_block3_3_conv (Conv2D)    (None, 28, 28, 256)  16640       conv2_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_out (Add)          (None, 28, 28, 256)  0           max_pooling2d_6[0][0]            \n                                                                 conv2_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_preact_bn (BatchNo (None, 28, 28, 256)  1024        conv2_block3_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block1_preact_relu (Activ (None, 28, 28, 256)  0           conv3_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block1_1_conv (Conv2D)    (None, 28, 28, 128)  32768       conv3_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block1_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_1_relu (Activation (None, 28, 28, 128)  0           conv3_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block1_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block1_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_2_relu (Activation (None, 28, 28, 128)  0           conv3_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block1_0_conv (Conv2D)    (None, 28, 28, 512)  131584      conv3_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block1_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_out (Add)          (None, 28, 28, 512)  0           conv3_block1_0_conv[0][0]        \n                                                                 conv3_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block2_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block2_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block2_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_1_relu (Activation (None, 28, 28, 128)  0           conv3_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block2_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block2_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_2_relu (Activation (None, 28, 28, 128)  0           conv3_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block2_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_out (Add)          (None, 28, 28, 512)  0           conv3_block1_out[0][0]           \n                                                                 conv3_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block3_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block3_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block3_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_1_relu (Activation (None, 28, 28, 128)  0           conv3_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block3_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block3_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_2_relu (Activation (None, 28, 28, 128)  0           conv3_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block3_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_out (Add)          (None, 28, 28, 512)  0           conv3_block2_out[0][0]           \n                                                                 conv3_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block3_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block4_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block4_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block4_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block4_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block4_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block4_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_1_relu (Activation (None, 28, 28, 128)  0           conv3_block4_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block4_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block4_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block4_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block4_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block4_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_2_relu (Activation (None, 28, 28, 128)  0           conv3_block4_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block4_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block4_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_out (Add)          (None, 28, 28, 512)  0           conv3_block3_out[0][0]           \n                                                                 conv3_block4_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block4_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block5_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block5_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block5_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block5_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block5_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block5_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_1_relu (Activation (None, 28, 28, 128)  0           conv3_block5_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block5_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block5_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block5_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block5_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block5_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_2_relu (Activation (None, 28, 28, 128)  0           conv3_block5_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block5_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block5_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_out (Add)          (None, 28, 28, 512)  0           conv3_block4_out[0][0]           \n                                                                 conv3_block5_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block5_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block6_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block6_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block6_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block6_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block6_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block6_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_1_relu (Activation (None, 28, 28, 128)  0           conv3_block6_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block6_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block6_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block6_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block6_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block6_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_2_relu (Activation (None, 28, 28, 128)  0           conv3_block6_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block6_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block6_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_out (Add)          (None, 28, 28, 512)  0           conv3_block5_out[0][0]           \n                                                                 conv3_block6_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block6_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block7_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block7_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block7_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block7_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block7_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block7_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_1_relu (Activation (None, 28, 28, 128)  0           conv3_block7_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block7_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block7_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block7_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block7_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block7_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_2_relu (Activation (None, 28, 28, 128)  0           conv3_block7_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block7_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block7_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_out (Add)          (None, 28, 28, 512)  0           conv3_block6_out[0][0]           \n                                                                 conv3_block7_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block7_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block8_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block8_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block8_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block8_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block8_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block8_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_1_relu (Activation (None, 28, 28, 128)  0           conv3_block8_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block8_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block8_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_2_conv (Conv2D)    (None, 14, 14, 128)  147456      conv3_block8_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block8_2_bn (BatchNormali (None, 14, 14, 128)  512         conv3_block8_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_2_relu (Activation (None, 14, 14, 128)  0           conv3_block8_2_bn[0][0]          \n__________________________________________________________________________________________________\nmax_pooling2d_7 (MaxPooling2D)  (None, 14, 14, 512)  0           conv3_block7_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block8_3_conv (Conv2D)    (None, 14, 14, 512)  66048       conv3_block8_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_out (Add)          (None, 14, 14, 512)  0           max_pooling2d_7[0][0]            \n                                                                 conv3_block8_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_preact_bn (BatchNo (None, 14, 14, 512)  2048        conv3_block8_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block1_preact_relu (Activ (None, 14, 14, 512)  0           conv4_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block1_1_conv (Conv2D)    (None, 14, 14, 256)  131072      conv4_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block1_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_1_relu (Activation (None, 14, 14, 256)  0           conv4_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block1_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block1_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_2_relu (Activation (None, 14, 14, 256)  0           conv4_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block1_0_conv (Conv2D)    (None, 14, 14, 1024) 525312      conv4_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block1_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_out (Add)          (None, 14, 14, 1024) 0           conv4_block1_0_conv[0][0]        \n                                                                 conv4_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block2_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block2_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block2_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_1_relu (Activation (None, 14, 14, 256)  0           conv4_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block2_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block2_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_2_relu (Activation (None, 14, 14, 256)  0           conv4_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block2_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_out (Add)          (None, 14, 14, 1024) 0           conv4_block1_out[0][0]           \n                                                                 conv4_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block3_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block3_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block3_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_1_relu (Activation (None, 14, 14, 256)  0           conv4_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block3_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block3_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_2_relu (Activation (None, 14, 14, 256)  0           conv4_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block3_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_out (Add)          (None, 14, 14, 1024) 0           conv4_block2_out[0][0]           \n                                                                 conv4_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block3_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block4_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block4_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block4_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block4_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block4_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block4_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_1_relu (Activation (None, 14, 14, 256)  0           conv4_block4_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block4_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block4_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block4_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block4_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block4_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_2_relu (Activation (None, 14, 14, 256)  0           conv4_block4_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block4_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block4_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_out (Add)          (None, 14, 14, 1024) 0           conv4_block3_out[0][0]           \n                                                                 conv4_block4_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block4_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block5_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block5_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block5_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block5_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block5_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block5_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_1_relu (Activation (None, 14, 14, 256)  0           conv4_block5_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block5_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block5_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block5_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block5_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block5_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_2_relu (Activation (None, 14, 14, 256)  0           conv4_block5_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block5_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block5_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_out (Add)          (None, 14, 14, 1024) 0           conv4_block4_out[0][0]           \n                                                                 conv4_block5_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block5_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block6_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block6_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block6_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block6_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block6_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block6_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_1_relu (Activation (None, 14, 14, 256)  0           conv4_block6_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block6_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block6_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block6_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block6_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block6_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_2_relu (Activation (None, 14, 14, 256)  0           conv4_block6_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block6_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block6_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_out (Add)          (None, 14, 14, 1024) 0           conv4_block5_out[0][0]           \n                                                                 conv4_block6_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block6_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block7_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block7_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block7_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block7_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block7_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block7_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_1_relu (Activation (None, 14, 14, 256)  0           conv4_block7_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block7_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block7_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block7_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block7_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block7_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_2_relu (Activation (None, 14, 14, 256)  0           conv4_block7_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block7_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block7_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_out (Add)          (None, 14, 14, 1024) 0           conv4_block6_out[0][0]           \n                                                                 conv4_block7_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block7_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block8_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block8_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block8_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block8_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block8_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block8_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_1_relu (Activation (None, 14, 14, 256)  0           conv4_block8_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block8_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block8_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block8_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block8_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block8_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_2_relu (Activation (None, 14, 14, 256)  0           conv4_block8_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block8_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block8_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_out (Add)          (None, 14, 14, 1024) 0           conv4_block7_out[0][0]           \n                                                                 conv4_block8_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block8_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block9_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block9_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block9_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block9_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block9_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block9_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_1_relu (Activation (None, 14, 14, 256)  0           conv4_block9_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block9_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block9_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block9_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block9_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block9_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_2_relu (Activation (None, 14, 14, 256)  0           conv4_block9_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block9_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block9_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_out (Add)          (None, 14, 14, 1024) 0           conv4_block8_out[0][0]           \n                                                                 conv4_block9_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block10_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block9_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block10_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block10_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block10_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block10_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block10_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block10_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block10_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block10_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block10_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block10_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block10_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block10_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block10_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block10_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block10_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_out (Add)         (None, 14, 14, 1024) 0           conv4_block9_out[0][0]           \n                                                                 conv4_block10_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block10_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block11_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block11_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block11_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block11_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block11_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block11_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block11_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block11_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block11_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block11_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block11_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block11_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block11_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block11_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block11_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_out (Add)         (None, 14, 14, 1024) 0           conv4_block10_out[0][0]          \n                                                                 conv4_block11_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block11_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block12_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block12_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block12_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block12_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block12_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block12_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block12_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block12_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block12_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block12_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block12_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block12_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block12_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block12_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block12_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_out (Add)         (None, 14, 14, 1024) 0           conv4_block11_out[0][0]          \n                                                                 conv4_block12_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block12_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block13_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block13_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block13_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block13_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block13_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block13_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block13_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block13_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block13_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block13_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block13_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block13_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block13_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block13_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block13_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_out (Add)         (None, 14, 14, 1024) 0           conv4_block12_out[0][0]          \n                                                                 conv4_block13_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block13_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block14_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block14_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block14_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block14_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block14_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block14_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block14_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block14_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block14_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block14_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block14_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block14_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block14_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block14_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block14_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_out (Add)         (None, 14, 14, 1024) 0           conv4_block13_out[0][0]          \n                                                                 conv4_block14_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block14_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block15_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block15_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block15_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block15_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block15_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block15_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block15_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block15_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block15_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block15_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block15_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block15_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block15_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block15_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block15_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_out (Add)         (None, 14, 14, 1024) 0           conv4_block14_out[0][0]          \n                                                                 conv4_block15_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block15_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block16_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block16_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block16_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block16_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block16_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block16_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block16_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block16_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block16_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block16_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block16_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block16_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block16_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block16_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block16_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_out (Add)         (None, 14, 14, 1024) 0           conv4_block15_out[0][0]          \n                                                                 conv4_block16_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block16_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block17_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block17_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block17_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block17_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block17_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block17_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block17_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block17_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block17_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block17_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block17_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block17_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block17_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block17_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block17_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_out (Add)         (None, 14, 14, 1024) 0           conv4_block16_out[0][0]          \n                                                                 conv4_block17_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block17_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block18_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block18_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block18_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block18_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block18_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block18_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block18_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block18_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block18_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block18_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block18_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block18_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block18_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block18_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block18_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_out (Add)         (None, 14, 14, 1024) 0           conv4_block17_out[0][0]          \n                                                                 conv4_block18_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block18_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block19_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block19_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block19_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block19_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block19_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block19_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block19_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block19_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block19_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block19_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block19_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block19_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block19_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block19_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block19_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_out (Add)         (None, 14, 14, 1024) 0           conv4_block18_out[0][0]          \n                                                                 conv4_block19_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block19_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block20_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block20_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block20_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block20_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block20_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block20_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block20_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block20_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block20_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block20_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block20_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block20_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block20_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block20_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block20_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_out (Add)         (None, 14, 14, 1024) 0           conv4_block19_out[0][0]          \n                                                                 conv4_block20_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block20_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block21_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block21_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block21_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block21_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block21_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block21_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block21_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block21_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block21_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block21_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block21_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block21_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block21_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block21_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block21_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_out (Add)         (None, 14, 14, 1024) 0           conv4_block20_out[0][0]          \n                                                                 conv4_block21_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block21_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block22_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block22_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block22_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block22_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block22_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block22_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block22_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block22_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block22_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block22_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block22_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block22_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block22_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block22_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block22_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_out (Add)         (None, 14, 14, 1024) 0           conv4_block21_out[0][0]          \n                                                                 conv4_block22_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block22_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block23_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block23_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block23_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block23_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block23_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block23_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block23_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block23_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block23_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block23_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block23_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block23_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block23_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block23_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block23_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_out (Add)         (None, 14, 14, 1024) 0           conv4_block22_out[0][0]          \n                                                                 conv4_block23_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block23_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block24_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block24_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block24_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block24_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block24_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block24_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block24_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block24_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block24_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block24_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block24_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block24_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block24_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block24_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block24_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_out (Add)         (None, 14, 14, 1024) 0           conv4_block23_out[0][0]          \n                                                                 conv4_block24_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block24_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block25_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block25_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block25_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block25_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block25_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block25_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block25_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block25_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block25_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block25_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block25_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block25_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block25_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block25_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block25_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_out (Add)         (None, 14, 14, 1024) 0           conv4_block24_out[0][0]          \n                                                                 conv4_block25_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block25_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block26_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block26_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block26_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block26_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block26_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block26_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block26_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block26_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block26_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block26_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block26_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block26_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block26_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block26_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block26_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_out (Add)         (None, 14, 14, 1024) 0           conv4_block25_out[0][0]          \n                                                                 conv4_block26_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block26_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block27_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block27_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block27_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block27_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block27_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block27_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block27_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block27_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block27_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block27_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block27_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block27_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block27_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block27_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block27_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_out (Add)         (None, 14, 14, 1024) 0           conv4_block26_out[0][0]          \n                                                                 conv4_block27_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block27_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block28_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block28_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block28_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block28_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block28_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block28_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block28_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block28_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block28_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block28_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block28_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block28_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block28_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block28_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block28_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_out (Add)         (None, 14, 14, 1024) 0           conv4_block27_out[0][0]          \n                                                                 conv4_block28_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block28_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block29_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block29_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block29_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block29_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block29_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block29_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block29_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block29_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block29_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block29_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block29_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block29_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block29_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block29_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block29_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_out (Add)         (None, 14, 14, 1024) 0           conv4_block28_out[0][0]          \n                                                                 conv4_block29_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block29_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block30_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block30_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block30_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block30_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block30_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block30_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block30_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block30_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block30_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block30_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block30_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block30_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block30_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block30_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block30_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_out (Add)         (None, 14, 14, 1024) 0           conv4_block29_out[0][0]          \n                                                                 conv4_block30_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block30_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block31_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block31_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block31_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block31_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block31_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block31_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block31_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block31_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block31_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block31_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block31_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block31_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block31_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block31_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block31_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_out (Add)         (None, 14, 14, 1024) 0           conv4_block30_out[0][0]          \n                                                                 conv4_block31_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block31_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block32_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block32_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block32_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block32_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block32_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block32_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block32_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block32_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block32_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block32_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block32_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block32_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block32_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block32_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block32_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_out (Add)         (None, 14, 14, 1024) 0           conv4_block31_out[0][0]          \n                                                                 conv4_block32_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block32_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block33_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block33_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block33_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block33_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block33_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block33_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block33_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block33_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block33_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block33_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block33_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block33_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block33_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block33_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block33_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_out (Add)         (None, 14, 14, 1024) 0           conv4_block32_out[0][0]          \n                                                                 conv4_block33_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block33_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block34_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block34_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block34_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block34_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block34_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block34_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block34_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block34_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block34_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block34_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block34_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block34_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block34_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block34_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block34_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_out (Add)         (None, 14, 14, 1024) 0           conv4_block33_out[0][0]          \n                                                                 conv4_block34_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block34_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block35_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block35_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block35_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block35_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block35_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block35_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block35_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block35_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block35_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block35_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block35_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block35_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block35_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block35_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block35_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_out (Add)         (None, 14, 14, 1024) 0           conv4_block34_out[0][0]          \n                                                                 conv4_block35_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block35_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block36_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block36_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block36_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block36_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block36_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block36_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block36_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block36_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block36_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_2_conv (Conv2D)   (None, 7, 7, 256)    589824      conv4_block36_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block36_2_bn (BatchNormal (None, 7, 7, 256)    1024        conv4_block36_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_2_relu (Activatio (None, 7, 7, 256)    0           conv4_block36_2_bn[0][0]         \n__________________________________________________________________________________________________\nmax_pooling2d_8 (MaxPooling2D)  (None, 7, 7, 1024)   0           conv4_block35_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block36_3_conv (Conv2D)   (None, 7, 7, 1024)   263168      conv4_block36_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_out (Add)         (None, 7, 7, 1024)   0           max_pooling2d_8[0][0]            \n                                                                 conv4_block36_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv5_block1_preact_bn (BatchNo (None, 7, 7, 1024)   4096        conv4_block36_out[0][0]          \n__________________________________________________________________________________________________\nconv5_block1_preact_relu (Activ (None, 7, 7, 1024)   0           conv5_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv5_block1_1_conv (Conv2D)    (None, 7, 7, 512)    524288      conv5_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block1_1_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_1_relu (Activation (None, 7, 7, 512)    0           conv5_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block1_2_pad (ZeroPadding (None, 9, 9, 512)    0           conv5_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_2_conv (Conv2D)    (None, 7, 7, 512)    2359296     conv5_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv5_block1_2_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_2_relu (Activation (None, 7, 7, 512)    0           conv5_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block1_0_conv (Conv2D)    (None, 7, 7, 2048)   2099200     conv5_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block1_3_conv (Conv2D)    (None, 7, 7, 2048)   1050624     conv5_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_out (Add)          (None, 7, 7, 2048)   0           conv5_block1_0_conv[0][0]        \n                                                                 conv5_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_preact_bn (BatchNo (None, 7, 7, 2048)   8192        conv5_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv5_block2_preact_relu (Activ (None, 7, 7, 2048)   0           conv5_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv5_block2_1_conv (Conv2D)    (None, 7, 7, 512)    1048576     conv5_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block2_1_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_1_relu (Activation (None, 7, 7, 512)    0           conv5_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block2_2_pad (ZeroPadding (None, 9, 9, 512)    0           conv5_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_2_conv (Conv2D)    (None, 7, 7, 512)    2359296     conv5_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv5_block2_2_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_2_relu (Activation (None, 7, 7, 512)    0           conv5_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block2_3_conv (Conv2D)    (None, 7, 7, 2048)   1050624     conv5_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_out (Add)          (None, 7, 7, 2048)   0           conv5_block1_out[0][0]           \n                                                                 conv5_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_preact_bn (BatchNo (None, 7, 7, 2048)   8192        conv5_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv5_block3_preact_relu (Activ (None, 7, 7, 2048)   0           conv5_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv5_block3_1_conv (Conv2D)    (None, 7, 7, 512)    1048576     conv5_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block3_1_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_1_relu (Activation (None, 7, 7, 512)    0           conv5_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block3_2_pad (ZeroPadding (None, 9, 9, 512)    0           conv5_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_2_conv (Conv2D)    (None, 7, 7, 512)    2359296     conv5_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv5_block3_2_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_2_relu (Activation (None, 7, 7, 512)    0           conv5_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block3_3_conv (Conv2D)    (None, 7, 7, 2048)   1050624     conv5_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_out (Add)          (None, 7, 7, 2048)   0           conv5_block2_out[0][0]           \n                                                                 conv5_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\npost_bn (BatchNormalization)    (None, 7, 7, 2048)   8192        conv5_block3_out[0][0]           \n__________________________________________________________________________________________________\npost_relu (Activation)          (None, 7, 7, 2048)   0           post_bn[0][0]                    \n==================================================================================================\nTotal params: 58,331,648\nTrainable params: 0\nNon-trainable params: 58,331,648\n__________________________________________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"markdown","source":"*Get the last layer*","metadata":{}},{"cell_type":"code","source":"last_layer = pre_trained_model.get_layer('post_relu')\nlast_output = last_layer.output","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:41.993593Z","iopub.execute_input":"2022-09-16T13:28:41.993948Z","iopub.status.idle":"2022-09-16T13:28:42.003523Z","shell.execute_reply.started":"2022-09-16T13:28:41.993911Z","shell.execute_reply":"2022-09-16T13:28:42.002085Z"},"trusted":true},"execution_count":21,"outputs":[]},{"cell_type":"markdown","source":"*Use the entire previus model and add few layers to train*","metadata":{}},{"cell_type":"code","source":"from tensorflow.keras.layers import Flatten","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.005452Z","iopub.execute_input":"2022-09-16T13:28:42.006270Z","iopub.status.idle":"2022-09-16T13:28:42.015542Z","shell.execute_reply.started":"2022-09-16T13:28:42.006222Z","shell.execute_reply":"2022-09-16T13:28:42.014069Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"x = Flatten()(last_output)\nx = Dense(units=128, activation = 'relu')(x)\nx = Dropout(0.2)(x)\nx = Dense(units=50, activation = 'relu')(x)\nx = Dropout(0.2)(x)\nx = Dense(units=3, activation = 'softmax')(x)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.017621Z","iopub.execute_input":"2022-09-16T13:28:42.018150Z","iopub.status.idle":"2022-09-16T13:28:42.065021Z","shell.execute_reply.started":"2022-09-16T13:28:42.018101Z","shell.execute_reply":"2022-09-16T13:28:42.063781Z"},"trusted":true},"execution_count":23,"outputs":[]},{"cell_type":"code","source":"import tensorflow as tf","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.071462Z","iopub.execute_input":"2022-09-16T13:28:42.071803Z","iopub.status.idle":"2022-09-16T13:28:42.080288Z","shell.execute_reply.started":"2022-09-16T13:28:42.071774Z","shell.execute_reply":"2022-09-16T13:28:42.078384Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"code","source":"model_tl = tf.keras.Model(pre_trained_model.input, x)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.082177Z","iopub.execute_input":"2022-09-16T13:28:42.083403Z","iopub.status.idle":"2022-09-16T13:28:42.128554Z","shell.execute_reply.started":"2022-09-16T13:28:42.083181Z","shell.execute_reply":"2022-09-16T13:28:42.127204Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"code","source":"model_tl.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.130454Z","iopub.execute_input":"2022-09-16T13:28:42.131194Z","iopub.status.idle":"2022-09-16T13:28:42.153021Z","shell.execute_reply.started":"2022-09-16T13:28:42.131147Z","shell.execute_reply":"2022-09-16T13:28:42.151569Z"},"trusted":true},"execution_count":26,"outputs":[]},{"cell_type":"markdown","source":"See the new trainable parameters of the model","metadata":{}},{"cell_type":"code","source":"model_tl.summary()","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.156814Z","iopub.execute_input":"2022-09-16T13:28:42.158206Z","iopub.status.idle":"2022-09-16T13:28:42.246140Z","shell.execute_reply.started":"2022-09-16T13:28:42.158150Z","shell.execute_reply":"2022-09-16T13:28:42.244933Z"},"trusted":true},"execution_count":27,"outputs":[{"name":"stdout","text":"Model: \"model\"\n__________________________________________________________________________________________________\nLayer (type)                    Output Shape         Param #     Connected to                     \n==================================================================================================\ninput_1 (InputLayer)            [(None, 224, 224, 3) 0                                            \n__________________________________________________________________________________________________\nconv1_pad (ZeroPadding2D)       (None, 230, 230, 3)  0           input_1[0][0]                    \n__________________________________________________________________________________________________\nconv1_conv (Conv2D)             (None, 112, 112, 64) 9472        conv1_pad[0][0]                  \n__________________________________________________________________________________________________\npool1_pad (ZeroPadding2D)       (None, 114, 114, 64) 0           conv1_conv[0][0]                 \n__________________________________________________________________________________________________\npool1_pool (MaxPooling2D)       (None, 56, 56, 64)   0           pool1_pad[0][0]                  \n__________________________________________________________________________________________________\nconv2_block1_preact_bn (BatchNo (None, 56, 56, 64)   256         pool1_pool[0][0]                 \n__________________________________________________________________________________________________\nconv2_block1_preact_relu (Activ (None, 56, 56, 64)   0           conv2_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv2_block1_1_conv (Conv2D)    (None, 56, 56, 64)   4096        conv2_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block1_1_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_1_relu (Activation (None, 56, 56, 64)   0           conv2_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block1_2_pad (ZeroPadding (None, 58, 58, 64)   0           conv2_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_2_conv (Conv2D)    (None, 56, 56, 64)   36864       conv2_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv2_block1_2_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_2_relu (Activation (None, 56, 56, 64)   0           conv2_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block1_0_conv (Conv2D)    (None, 56, 56, 256)  16640       conv2_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block1_3_conv (Conv2D)    (None, 56, 56, 256)  16640       conv2_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block1_out (Add)          (None, 56, 56, 256)  0           conv2_block1_0_conv[0][0]        \n                                                                 conv2_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_preact_bn (BatchNo (None, 56, 56, 256)  1024        conv2_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv2_block2_preact_relu (Activ (None, 56, 56, 256)  0           conv2_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv2_block2_1_conv (Conv2D)    (None, 56, 56, 64)   16384       conv2_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block2_1_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_1_relu (Activation (None, 56, 56, 64)   0           conv2_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block2_2_pad (ZeroPadding (None, 58, 58, 64)   0           conv2_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_2_conv (Conv2D)    (None, 56, 56, 64)   36864       conv2_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv2_block2_2_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_2_relu (Activation (None, 56, 56, 64)   0           conv2_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block2_3_conv (Conv2D)    (None, 56, 56, 256)  16640       conv2_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block2_out (Add)          (None, 56, 56, 256)  0           conv2_block1_out[0][0]           \n                                                                 conv2_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_preact_bn (BatchNo (None, 56, 56, 256)  1024        conv2_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv2_block3_preact_relu (Activ (None, 56, 56, 256)  0           conv2_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv2_block3_1_conv (Conv2D)    (None, 56, 56, 64)   16384       conv2_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv2_block3_1_bn (BatchNormali (None, 56, 56, 64)   256         conv2_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_1_relu (Activation (None, 56, 56, 64)   0           conv2_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv2_block3_2_pad (ZeroPadding (None, 58, 58, 64)   0           conv2_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_2_conv (Conv2D)    (None, 28, 28, 64)   36864       conv2_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv2_block3_2_bn (BatchNormali (None, 28, 28, 64)   256         conv2_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_2_relu (Activation (None, 28, 28, 64)   0           conv2_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nmax_pooling2d_6 (MaxPooling2D)  (None, 28, 28, 256)  0           conv2_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv2_block3_3_conv (Conv2D)    (None, 28, 28, 256)  16640       conv2_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv2_block3_out (Add)          (None, 28, 28, 256)  0           max_pooling2d_6[0][0]            \n                                                                 conv2_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_preact_bn (BatchNo (None, 28, 28, 256)  1024        conv2_block3_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block1_preact_relu (Activ (None, 28, 28, 256)  0           conv3_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block1_1_conv (Conv2D)    (None, 28, 28, 128)  32768       conv3_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block1_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_1_relu (Activation (None, 28, 28, 128)  0           conv3_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block1_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block1_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_2_relu (Activation (None, 28, 28, 128)  0           conv3_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block1_0_conv (Conv2D)    (None, 28, 28, 512)  131584      conv3_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block1_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block1_out (Add)          (None, 28, 28, 512)  0           conv3_block1_0_conv[0][0]        \n                                                                 conv3_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block2_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block2_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block2_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_1_relu (Activation (None, 28, 28, 128)  0           conv3_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block2_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block2_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_2_relu (Activation (None, 28, 28, 128)  0           conv3_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block2_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block2_out (Add)          (None, 28, 28, 512)  0           conv3_block1_out[0][0]           \n                                                                 conv3_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block3_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block3_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block3_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_1_relu (Activation (None, 28, 28, 128)  0           conv3_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block3_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block3_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_2_relu (Activation (None, 28, 28, 128)  0           conv3_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block3_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block3_out (Add)          (None, 28, 28, 512)  0           conv3_block2_out[0][0]           \n                                                                 conv3_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block3_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block4_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block4_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block4_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block4_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block4_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block4_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_1_relu (Activation (None, 28, 28, 128)  0           conv3_block4_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block4_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block4_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block4_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block4_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block4_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_2_relu (Activation (None, 28, 28, 128)  0           conv3_block4_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block4_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block4_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block4_out (Add)          (None, 28, 28, 512)  0           conv3_block3_out[0][0]           \n                                                                 conv3_block4_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block4_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block5_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block5_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block5_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block5_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block5_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block5_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_1_relu (Activation (None, 28, 28, 128)  0           conv3_block5_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block5_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block5_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block5_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block5_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block5_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_2_relu (Activation (None, 28, 28, 128)  0           conv3_block5_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block5_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block5_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block5_out (Add)          (None, 28, 28, 512)  0           conv3_block4_out[0][0]           \n                                                                 conv3_block5_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block5_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block6_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block6_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block6_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block6_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block6_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block6_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_1_relu (Activation (None, 28, 28, 128)  0           conv3_block6_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block6_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block6_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block6_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block6_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block6_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_2_relu (Activation (None, 28, 28, 128)  0           conv3_block6_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block6_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block6_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block6_out (Add)          (None, 28, 28, 512)  0           conv3_block5_out[0][0]           \n                                                                 conv3_block6_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block6_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block7_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block7_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block7_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block7_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block7_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block7_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_1_relu (Activation (None, 28, 28, 128)  0           conv3_block7_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block7_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block7_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_2_conv (Conv2D)    (None, 28, 28, 128)  147456      conv3_block7_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block7_2_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block7_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_2_relu (Activation (None, 28, 28, 128)  0           conv3_block7_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block7_3_conv (Conv2D)    (None, 28, 28, 512)  66048       conv3_block7_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block7_out (Add)          (None, 28, 28, 512)  0           conv3_block6_out[0][0]           \n                                                                 conv3_block7_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_preact_bn (BatchNo (None, 28, 28, 512)  2048        conv3_block7_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block8_preact_relu (Activ (None, 28, 28, 512)  0           conv3_block8_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv3_block8_1_conv (Conv2D)    (None, 28, 28, 128)  65536       conv3_block8_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv3_block8_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block8_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_1_relu (Activation (None, 28, 28, 128)  0           conv3_block8_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv3_block8_2_pad (ZeroPadding (None, 30, 30, 128)  0           conv3_block8_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_2_conv (Conv2D)    (None, 14, 14, 128)  147456      conv3_block8_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv3_block8_2_bn (BatchNormali (None, 14, 14, 128)  512         conv3_block8_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_2_relu (Activation (None, 14, 14, 128)  0           conv3_block8_2_bn[0][0]          \n__________________________________________________________________________________________________\nmax_pooling2d_7 (MaxPooling2D)  (None, 14, 14, 512)  0           conv3_block7_out[0][0]           \n__________________________________________________________________________________________________\nconv3_block8_3_conv (Conv2D)    (None, 14, 14, 512)  66048       conv3_block8_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv3_block8_out (Add)          (None, 14, 14, 512)  0           max_pooling2d_7[0][0]            \n                                                                 conv3_block8_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_preact_bn (BatchNo (None, 14, 14, 512)  2048        conv3_block8_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block1_preact_relu (Activ (None, 14, 14, 512)  0           conv4_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block1_1_conv (Conv2D)    (None, 14, 14, 256)  131072      conv4_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block1_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_1_relu (Activation (None, 14, 14, 256)  0           conv4_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block1_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block1_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_2_relu (Activation (None, 14, 14, 256)  0           conv4_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block1_0_conv (Conv2D)    (None, 14, 14, 1024) 525312      conv4_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block1_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block1_out (Add)          (None, 14, 14, 1024) 0           conv4_block1_0_conv[0][0]        \n                                                                 conv4_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block2_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block2_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block2_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_1_relu (Activation (None, 14, 14, 256)  0           conv4_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block2_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block2_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_2_relu (Activation (None, 14, 14, 256)  0           conv4_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block2_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block2_out (Add)          (None, 14, 14, 1024) 0           conv4_block1_out[0][0]           \n                                                                 conv4_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block3_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block3_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block3_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_1_relu (Activation (None, 14, 14, 256)  0           conv4_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block3_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block3_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_2_relu (Activation (None, 14, 14, 256)  0           conv4_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block3_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block3_out (Add)          (None, 14, 14, 1024) 0           conv4_block2_out[0][0]           \n                                                                 conv4_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block3_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block4_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block4_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block4_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block4_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block4_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block4_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_1_relu (Activation (None, 14, 14, 256)  0           conv4_block4_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block4_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block4_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block4_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block4_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block4_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_2_relu (Activation (None, 14, 14, 256)  0           conv4_block4_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block4_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block4_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block4_out (Add)          (None, 14, 14, 1024) 0           conv4_block3_out[0][0]           \n                                                                 conv4_block4_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block4_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block5_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block5_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block5_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block5_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block5_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block5_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_1_relu (Activation (None, 14, 14, 256)  0           conv4_block5_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block5_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block5_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block5_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block5_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block5_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_2_relu (Activation (None, 14, 14, 256)  0           conv4_block5_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block5_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block5_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block5_out (Add)          (None, 14, 14, 1024) 0           conv4_block4_out[0][0]           \n                                                                 conv4_block5_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block5_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block6_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block6_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block6_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block6_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block6_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block6_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_1_relu (Activation (None, 14, 14, 256)  0           conv4_block6_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block6_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block6_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block6_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block6_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block6_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_2_relu (Activation (None, 14, 14, 256)  0           conv4_block6_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block6_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block6_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block6_out (Add)          (None, 14, 14, 1024) 0           conv4_block5_out[0][0]           \n                                                                 conv4_block6_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block6_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block7_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block7_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block7_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block7_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block7_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block7_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_1_relu (Activation (None, 14, 14, 256)  0           conv4_block7_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block7_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block7_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block7_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block7_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block7_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_2_relu (Activation (None, 14, 14, 256)  0           conv4_block7_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block7_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block7_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block7_out (Add)          (None, 14, 14, 1024) 0           conv4_block6_out[0][0]           \n                                                                 conv4_block7_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block7_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block8_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block8_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block8_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block8_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block8_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block8_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_1_relu (Activation (None, 14, 14, 256)  0           conv4_block8_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block8_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block8_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block8_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block8_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block8_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_2_relu (Activation (None, 14, 14, 256)  0           conv4_block8_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block8_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block8_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block8_out (Add)          (None, 14, 14, 1024) 0           conv4_block7_out[0][0]           \n                                                                 conv4_block8_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_preact_bn (BatchNo (None, 14, 14, 1024) 4096        conv4_block8_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block9_preact_relu (Activ (None, 14, 14, 1024) 0           conv4_block9_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv4_block9_1_conv (Conv2D)    (None, 14, 14, 256)  262144      conv4_block9_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv4_block9_1_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block9_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_1_relu (Activation (None, 14, 14, 256)  0           conv4_block9_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block9_2_pad (ZeroPadding (None, 16, 16, 256)  0           conv4_block9_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_2_conv (Conv2D)    (None, 14, 14, 256)  589824      conv4_block9_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv4_block9_2_bn (BatchNormali (None, 14, 14, 256)  1024        conv4_block9_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_2_relu (Activation (None, 14, 14, 256)  0           conv4_block9_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv4_block9_3_conv (Conv2D)    (None, 14, 14, 1024) 263168      conv4_block9_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv4_block9_out (Add)          (None, 14, 14, 1024) 0           conv4_block8_out[0][0]           \n                                                                 conv4_block9_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv4_block10_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block9_out[0][0]           \n__________________________________________________________________________________________________\nconv4_block10_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block10_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block10_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block10_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block10_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block10_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block10_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block10_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block10_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block10_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block10_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block10_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block10_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block10_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block10_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block10_out (Add)         (None, 14, 14, 1024) 0           conv4_block9_out[0][0]           \n                                                                 conv4_block10_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block10_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block11_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block11_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block11_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block11_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block11_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block11_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block11_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block11_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block11_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block11_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block11_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block11_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block11_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block11_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block11_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block11_out (Add)         (None, 14, 14, 1024) 0           conv4_block10_out[0][0]          \n                                                                 conv4_block11_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block11_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block12_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block12_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block12_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block12_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block12_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block12_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block12_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block12_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block12_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block12_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block12_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block12_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block12_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block12_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block12_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block12_out (Add)         (None, 14, 14, 1024) 0           conv4_block11_out[0][0]          \n                                                                 conv4_block12_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block12_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block13_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block13_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block13_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block13_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block13_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block13_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block13_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block13_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block13_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block13_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block13_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block13_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block13_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block13_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block13_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block13_out (Add)         (None, 14, 14, 1024) 0           conv4_block12_out[0][0]          \n                                                                 conv4_block13_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block13_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block14_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block14_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block14_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block14_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block14_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block14_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block14_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block14_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block14_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block14_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block14_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block14_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block14_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block14_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block14_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block14_out (Add)         (None, 14, 14, 1024) 0           conv4_block13_out[0][0]          \n                                                                 conv4_block14_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block14_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block15_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block15_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block15_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block15_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block15_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block15_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block15_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block15_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block15_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block15_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block15_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block15_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block15_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block15_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block15_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block15_out (Add)         (None, 14, 14, 1024) 0           conv4_block14_out[0][0]          \n                                                                 conv4_block15_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block15_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block16_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block16_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block16_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block16_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block16_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block16_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block16_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block16_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block16_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block16_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block16_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block16_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block16_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block16_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block16_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block16_out (Add)         (None, 14, 14, 1024) 0           conv4_block15_out[0][0]          \n                                                                 conv4_block16_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block16_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block17_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block17_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block17_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block17_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block17_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block17_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block17_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block17_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block17_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block17_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block17_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block17_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block17_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block17_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block17_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block17_out (Add)         (None, 14, 14, 1024) 0           conv4_block16_out[0][0]          \n                                                                 conv4_block17_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block17_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block18_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block18_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block18_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block18_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block18_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block18_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block18_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block18_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block18_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block18_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block18_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block18_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block18_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block18_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block18_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block18_out (Add)         (None, 14, 14, 1024) 0           conv4_block17_out[0][0]          \n                                                                 conv4_block18_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block18_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block19_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block19_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block19_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block19_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block19_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block19_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block19_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block19_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block19_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block19_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block19_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block19_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block19_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block19_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block19_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block19_out (Add)         (None, 14, 14, 1024) 0           conv4_block18_out[0][0]          \n                                                                 conv4_block19_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block19_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block20_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block20_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block20_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block20_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block20_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block20_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block20_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block20_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block20_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block20_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block20_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block20_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block20_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block20_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block20_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block20_out (Add)         (None, 14, 14, 1024) 0           conv4_block19_out[0][0]          \n                                                                 conv4_block20_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block20_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block21_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block21_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block21_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block21_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block21_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block21_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block21_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block21_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block21_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block21_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block21_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block21_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block21_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block21_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block21_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block21_out (Add)         (None, 14, 14, 1024) 0           conv4_block20_out[0][0]          \n                                                                 conv4_block21_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block21_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block22_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block22_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block22_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block22_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block22_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block22_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block22_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block22_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block22_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block22_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block22_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block22_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block22_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block22_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block22_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block22_out (Add)         (None, 14, 14, 1024) 0           conv4_block21_out[0][0]          \n                                                                 conv4_block22_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block22_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block23_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block23_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block23_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block23_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block23_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block23_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block23_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block23_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block23_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block23_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block23_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block23_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block23_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block23_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block23_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block23_out (Add)         (None, 14, 14, 1024) 0           conv4_block22_out[0][0]          \n                                                                 conv4_block23_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block23_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block24_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block24_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block24_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block24_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block24_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block24_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block24_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block24_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block24_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block24_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block24_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block24_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block24_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block24_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block24_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block24_out (Add)         (None, 14, 14, 1024) 0           conv4_block23_out[0][0]          \n                                                                 conv4_block24_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block24_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block25_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block25_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block25_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block25_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block25_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block25_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block25_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block25_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block25_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block25_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block25_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block25_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block25_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block25_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block25_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block25_out (Add)         (None, 14, 14, 1024) 0           conv4_block24_out[0][0]          \n                                                                 conv4_block25_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block25_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block26_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block26_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block26_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block26_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block26_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block26_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block26_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block26_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block26_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block26_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block26_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block26_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block26_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block26_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block26_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block26_out (Add)         (None, 14, 14, 1024) 0           conv4_block25_out[0][0]          \n                                                                 conv4_block26_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block26_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block27_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block27_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block27_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block27_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block27_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block27_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block27_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block27_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block27_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block27_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block27_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block27_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block27_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block27_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block27_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block27_out (Add)         (None, 14, 14, 1024) 0           conv4_block26_out[0][0]          \n                                                                 conv4_block27_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block27_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block28_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block28_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block28_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block28_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block28_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block28_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block28_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block28_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block28_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block28_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block28_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block28_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block28_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block28_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block28_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block28_out (Add)         (None, 14, 14, 1024) 0           conv4_block27_out[0][0]          \n                                                                 conv4_block28_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block28_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block29_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block29_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block29_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block29_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block29_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block29_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block29_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block29_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block29_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block29_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block29_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block29_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block29_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block29_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block29_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block29_out (Add)         (None, 14, 14, 1024) 0           conv4_block28_out[0][0]          \n                                                                 conv4_block29_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block29_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block30_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block30_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block30_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block30_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block30_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block30_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block30_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block30_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block30_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block30_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block30_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block30_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block30_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block30_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block30_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block30_out (Add)         (None, 14, 14, 1024) 0           conv4_block29_out[0][0]          \n                                                                 conv4_block30_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block30_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block31_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block31_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block31_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block31_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block31_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block31_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block31_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block31_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block31_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block31_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block31_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block31_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block31_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block31_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block31_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block31_out (Add)         (None, 14, 14, 1024) 0           conv4_block30_out[0][0]          \n                                                                 conv4_block31_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block31_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block32_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block32_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block32_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block32_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block32_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block32_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block32_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block32_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block32_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block32_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block32_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block32_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block32_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block32_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block32_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block32_out (Add)         (None, 14, 14, 1024) 0           conv4_block31_out[0][0]          \n                                                                 conv4_block32_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block32_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block33_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block33_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block33_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block33_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block33_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block33_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block33_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block33_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block33_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block33_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block33_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block33_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block33_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block33_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block33_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block33_out (Add)         (None, 14, 14, 1024) 0           conv4_block32_out[0][0]          \n                                                                 conv4_block33_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block33_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block34_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block34_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block34_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block34_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block34_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block34_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block34_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block34_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block34_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block34_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block34_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block34_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block34_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block34_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block34_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block34_out (Add)         (None, 14, 14, 1024) 0           conv4_block33_out[0][0]          \n                                                                 conv4_block34_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block34_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block35_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block35_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block35_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block35_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block35_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block35_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block35_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block35_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block35_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_2_conv (Conv2D)   (None, 14, 14, 256)  589824      conv4_block35_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block35_2_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block35_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_2_relu (Activatio (None, 14, 14, 256)  0           conv4_block35_2_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block35_3_conv (Conv2D)   (None, 14, 14, 1024) 263168      conv4_block35_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block35_out (Add)         (None, 14, 14, 1024) 0           conv4_block34_out[0][0]          \n                                                                 conv4_block35_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_preact_bn (BatchN (None, 14, 14, 1024) 4096        conv4_block35_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block36_preact_relu (Acti (None, 14, 14, 1024) 0           conv4_block36_preact_bn[0][0]    \n__________________________________________________________________________________________________\nconv4_block36_1_conv (Conv2D)   (None, 14, 14, 256)  262144      conv4_block36_preact_relu[0][0]  \n__________________________________________________________________________________________________\nconv4_block36_1_bn (BatchNormal (None, 14, 14, 256)  1024        conv4_block36_1_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_1_relu (Activatio (None, 14, 14, 256)  0           conv4_block36_1_bn[0][0]         \n__________________________________________________________________________________________________\nconv4_block36_2_pad (ZeroPaddin (None, 16, 16, 256)  0           conv4_block36_1_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_2_conv (Conv2D)   (None, 7, 7, 256)    589824      conv4_block36_2_pad[0][0]        \n__________________________________________________________________________________________________\nconv4_block36_2_bn (BatchNormal (None, 7, 7, 256)    1024        conv4_block36_2_conv[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_2_relu (Activatio (None, 7, 7, 256)    0           conv4_block36_2_bn[0][0]         \n__________________________________________________________________________________________________\nmax_pooling2d_8 (MaxPooling2D)  (None, 7, 7, 1024)   0           conv4_block35_out[0][0]          \n__________________________________________________________________________________________________\nconv4_block36_3_conv (Conv2D)   (None, 7, 7, 1024)   263168      conv4_block36_2_relu[0][0]       \n__________________________________________________________________________________________________\nconv4_block36_out (Add)         (None, 7, 7, 1024)   0           max_pooling2d_8[0][0]            \n                                                                 conv4_block36_3_conv[0][0]       \n__________________________________________________________________________________________________\nconv5_block1_preact_bn (BatchNo (None, 7, 7, 1024)   4096        conv4_block36_out[0][0]          \n__________________________________________________________________________________________________\nconv5_block1_preact_relu (Activ (None, 7, 7, 1024)   0           conv5_block1_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv5_block1_1_conv (Conv2D)    (None, 7, 7, 512)    524288      conv5_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block1_1_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block1_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_1_relu (Activation (None, 7, 7, 512)    0           conv5_block1_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block1_2_pad (ZeroPadding (None, 9, 9, 512)    0           conv5_block1_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_2_conv (Conv2D)    (None, 7, 7, 512)    2359296     conv5_block1_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv5_block1_2_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block1_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_2_relu (Activation (None, 7, 7, 512)    0           conv5_block1_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block1_0_conv (Conv2D)    (None, 7, 7, 2048)   2099200     conv5_block1_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block1_3_conv (Conv2D)    (None, 7, 7, 2048)   1050624     conv5_block1_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block1_out (Add)          (None, 7, 7, 2048)   0           conv5_block1_0_conv[0][0]        \n                                                                 conv5_block1_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_preact_bn (BatchNo (None, 7, 7, 2048)   8192        conv5_block1_out[0][0]           \n__________________________________________________________________________________________________\nconv5_block2_preact_relu (Activ (None, 7, 7, 2048)   0           conv5_block2_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv5_block2_1_conv (Conv2D)    (None, 7, 7, 512)    1048576     conv5_block2_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block2_1_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block2_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_1_relu (Activation (None, 7, 7, 512)    0           conv5_block2_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block2_2_pad (ZeroPadding (None, 9, 9, 512)    0           conv5_block2_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_2_conv (Conv2D)    (None, 7, 7, 512)    2359296     conv5_block2_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv5_block2_2_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block2_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_2_relu (Activation (None, 7, 7, 512)    0           conv5_block2_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block2_3_conv (Conv2D)    (None, 7, 7, 2048)   1050624     conv5_block2_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block2_out (Add)          (None, 7, 7, 2048)   0           conv5_block1_out[0][0]           \n                                                                 conv5_block2_3_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_preact_bn (BatchNo (None, 7, 7, 2048)   8192        conv5_block2_out[0][0]           \n__________________________________________________________________________________________________\nconv5_block3_preact_relu (Activ (None, 7, 7, 2048)   0           conv5_block3_preact_bn[0][0]     \n__________________________________________________________________________________________________\nconv5_block3_1_conv (Conv2D)    (None, 7, 7, 512)    1048576     conv5_block3_preact_relu[0][0]   \n__________________________________________________________________________________________________\nconv5_block3_1_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block3_1_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_1_relu (Activation (None, 7, 7, 512)    0           conv5_block3_1_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block3_2_pad (ZeroPadding (None, 9, 9, 512)    0           conv5_block3_1_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_2_conv (Conv2D)    (None, 7, 7, 512)    2359296     conv5_block3_2_pad[0][0]         \n__________________________________________________________________________________________________\nconv5_block3_2_bn (BatchNormali (None, 7, 7, 512)    2048        conv5_block3_2_conv[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_2_relu (Activation (None, 7, 7, 512)    0           conv5_block3_2_bn[0][0]          \n__________________________________________________________________________________________________\nconv5_block3_3_conv (Conv2D)    (None, 7, 7, 2048)   1050624     conv5_block3_2_relu[0][0]        \n__________________________________________________________________________________________________\nconv5_block3_out (Add)          (None, 7, 7, 2048)   0           conv5_block2_out[0][0]           \n                                                                 conv5_block3_3_conv[0][0]        \n__________________________________________________________________________________________________\npost_bn (BatchNormalization)    (None, 7, 7, 2048)   8192        conv5_block3_out[0][0]           \n__________________________________________________________________________________________________\npost_relu (Activation)          (None, 7, 7, 2048)   0           post_bn[0][0]                    \n__________________________________________________________________________________________________\nflatten_2 (Flatten)             (None, 100352)       0           post_relu[0][0]                  \n__________________________________________________________________________________________________\ndense_3 (Dense)                 (None, 128)          12845184    flatten_2[0][0]                  \n__________________________________________________________________________________________________\ndropout_7 (Dropout)             (None, 128)          0           dense_3[0][0]                    \n__________________________________________________________________________________________________\ndense_4 (Dense)                 (None, 50)           6450        dropout_7[0][0]                  \n__________________________________________________________________________________________________\ndropout_8 (Dropout)             (None, 50)           0           dense_4[0][0]                    \n__________________________________________________________________________________________________\ndense_5 (Dense)                 (None, 3)            153         dropout_8[0][0]                  \n==================================================================================================\nTotal params: 71,183,435\nTrainable params: 12,851,787\nNon-trainable params: 58,331,648\n__________________________________________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"model_tl.fit(train_data_resize, epochs = 15, validation_data = val_data_resize, verbose = 2)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:28:42.248232Z","iopub.execute_input":"2022-09-16T13:28:42.249045Z","iopub.status.idle":"2022-09-16T13:32:20.627605Z","shell.execute_reply.started":"2022-09-16T13:28:42.249002Z","shell.execute_reply":"2022-09-16T13:32:20.626231Z"},"trusted":true},"execution_count":28,"outputs":[{"name":"stdout","text":"Epoch 1/15\n7/7 - 30s - loss: 10.1729 - accuracy: 0.5920 - val_loss: 3.4295 - val_accuracy: 0.7800\nEpoch 2/15\n7/7 - 13s - loss: 2.2507 - accuracy: 0.8308 - val_loss: 4.7401 - val_accuracy: 0.7800\nEpoch 3/15\n7/7 - 13s - loss: 2.4589 - accuracy: 0.8856 - val_loss: 1.9960 - val_accuracy: 0.8600\nEpoch 4/15\n7/7 - 12s - loss: 0.9133 - accuracy: 0.9254 - val_loss: 2.7696 - val_accuracy: 0.8800\nEpoch 5/15\n7/7 - 12s - loss: 0.8698 - accuracy: 0.9204 - val_loss: 1.2617 - val_accuracy: 0.9000\nEpoch 6/15\n7/7 - 13s - loss: 0.7058 - accuracy: 0.9353 - val_loss: 1.8978 - val_accuracy: 0.8600\nEpoch 7/15\n7/7 - 12s - loss: 0.1057 - accuracy: 0.9801 - val_loss: 1.7545 - val_accuracy: 0.9000\nEpoch 8/15\n7/7 - 14s - loss: 0.1070 - accuracy: 0.9801 - val_loss: 1.5525 - val_accuracy: 0.8800\nEpoch 9/15\n7/7 - 13s - loss: 0.1141 - accuracy: 0.9851 - val_loss: 1.3134 - val_accuracy: 0.8400\nEpoch 10/15\n7/7 - 12s - loss: 0.0361 - accuracy: 0.9851 - val_loss: 2.4963 - val_accuracy: 0.9200\nEpoch 11/15\n7/7 - 13s - loss: 0.2179 - accuracy: 0.9900 - val_loss: 2.6923 - val_accuracy: 0.9200\nEpoch 12/15\n7/7 - 13s - loss: 0.0557 - accuracy: 0.9950 - val_loss: 2.7733 - val_accuracy: 0.9000\nEpoch 13/15\n7/7 - 13s - loss: 0.1425 - accuracy: 0.9900 - val_loss: 3.5914 - val_accuracy: 0.9200\nEpoch 14/15\n7/7 - 12s - loss: 0.0497 - accuracy: 0.9950 - val_loss: 3.4612 - val_accuracy: 0.9000\nEpoch 15/15\n7/7 - 12s - loss: 0.1947 - accuracy: 0.9900 - val_loss: 2.7730 - val_accuracy: 0.9200\n","output_type":"stream"},{"execution_count":28,"output_type":"execute_result","data":{"text/plain":"<keras.callbacks.History at 0x7f1e5825bb50>"},"metadata":{}}]},{"cell_type":"markdown","source":"See if the algorithm was able to generalize and learn well","metadata":{}},{"cell_type":"code","source":"model_tl.evaluate(test_data_resize)","metadata":{"execution":{"iopub.status.busy":"2022-09-16T13:32:20.629987Z","iopub.execute_input":"2022-09-16T13:32:20.630446Z","iopub.status.idle":"2022-09-16T13:32:23.357568Z","shell.execute_reply.started":"2022-09-16T13:32:20.630389Z","shell.execute_reply":"2022-09-16T13:32:23.356320Z"},"trusted":true},"execution_count":29,"outputs":[{"name":"stdout","text":"3/3 [==============================] - 2s 629ms/step - loss: 1.0372 - accuracy: 0.9242\n","output_type":"stream"},{"execution_count":29,"output_type":"execute_result","data":{"text/plain":"[1.0372493267059326, 0.9242424368858337]"},"metadata":{}}]},{"cell_type":"markdown","source":"## Final conclusions\n\n1. With CNN architectures purposed was not possible to have acuraccies higher than 0.44 roughly. The CNN convolutes around this value. In some iterations, it is possible to get a higher accuracy, but, after the accuracy comes back to 0.44.\n\n2. The data is not enough to generalize the problem. It is recommended to increase the size of the data set.\n\n3. The data should be treated before the CNN model. For instance, it should be applied filters that could seem to give extra features for the classification.\n\n4. Only 5 epochs were done because more than this number, do not increase the accuracy in the model. However, the loss function result do increase. \n\n5. It was thought that I might be a overfitting issue, but with a simple CNN also was not able to get acuraccies higher than 0.5\n\n6. The best solution was using transfer learning. Transfer Learning uses another pretrained architecture and applies this into the new problem.","metadata":{}},{"cell_type":"markdown","source":"## Final results\n\nThe final model was able to get an acuraccy of 0.92 and a loss of 0.65, so this model can be deploy to predict COVID-19 disease based only in the X-ray","metadata":{}}]}