Siamese_Net.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Siamese Net.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyNXZK4VAodaiJYIHFWHc4KZ",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Agrover112/SiameseNet-Search/blob/main/Siamese_Net.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LPQGN2LI5j-8",
        "outputId": "f479c187-4eed-4155-f157-848ae445c270"
      },
      "source": [
        "\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import random\n",
        "\n",
        "\n",
        "\n",
        "print('TensorFlow version:', tf.__version__)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TensorFlow version: 2.3.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7G91LMKR6cce",
        "outputId": "74e255c3-ac72-4cb2-9cec-d8094dafbb9f"
      },
      "source": [
        "(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()\n",
        "print(\"X_train:\",x_train.shape,\"Y_train:\",y_train.shape,\"X_test:\",x_test.shape,\"Y_test:\",y_test.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 0s 0us/step\n",
            "X_train: (60000, 28, 28) Y_train: (60000,) X_test: (10000, 28, 28) Y_test: (10000,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R0WN1aOH8eCn",
        "outputId": "47d13826-15fc-4979-a092-3758d07b5d3b"
      },
      "source": [
        "x_train = np.reshape(x_train, (x_train.shape[0], 784))/255.\n",
        "x_test = np.reshape(x_test, (x_test.shape[0], 784))/255.\n",
        "print(x_train.shape,x_test.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 784) (10000, 784)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LPCcqFIU8lIP"
      },
      "source": [
        "def plot_triplets(examples):\n",
        "    plt.figure(figsize=(6, 2))\n",
        "    for i in range(3):\n",
        "        plt.subplot(1, 3, 1 + i)\n",
        "        plt.imshow(np.reshape(examples[i], (28, 28)), cmap='binary')\n",
        "        plt.xticks([])\n",
        "        plt.yticks([])\n",
        "    plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 133
        },
        "id": "9fucvXGH8m0u",
        "outputId": "e532eb01-fe24-4e96-b120-d4e0979d9658"
      },
      "source": [
        "\n",
        "plot_triplets([x_train[0], x_train[1], x_train[2]])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x144 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KwGXTgVe8q8-"
      },
      "source": [
        "def create_batch(batch_size=256):\n",
        "    x_anchors = np.zeros((batch_size, 784))\n",
        "    x_positives = np.zeros((batch_size, 784))\n",
        "    x_negatives = np.zeros((batch_size, 784))\n",
        "    \n",
        "    for i in range(0, batch_size):\n",
        "        # We need to find an anchor, a positive example and a negative example\n",
        "        random_index = random.randint(0, x_train.shape[0] - 1)\n",
        "        x_anchor = x_train[random_index]\n",
        "        y = y_train[random_index]\n",
        "        \n",
        "        indices_for_pos = np.squeeze(np.where(y_train == y))  #single d( Find and remove)\n",
        "        indices_for_neg = np.squeeze(np.where(y_train != y))\n",
        "        \n",
        "        x_positive = x_train[indices_for_pos[random.randint(0, len(indices_for_pos) - 1)]]\n",
        "        x_negative = x_train[indices_for_neg[random.randint(0, len(indices_for_neg) - 1)]]\n",
        "        \n",
        "        x_anchors[i] = x_anchor\n",
        "        x_positives[i] = x_positive\n",
        "        x_negatives[i] = x_negative\n",
        "        \n",
        "    return [x_anchors, x_positives, x_negatives]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 150
        },
        "id": "-HaviAsO8t9-",
        "outputId": "9fa79ae2-451d-4069-9783-9ce11c445791"
      },
      "source": [
        "examples = create_batch(1)\n",
        "print(examples[2].shape)\n",
        "plot_triplets(examples)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1, 784)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x144 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R82vpLtf-5cs"
      },
      "source": [
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WBhCT6Gt8vdG"
      },
      "source": [
        "emb_size = 64\n",
        "\n",
        "embedding_model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(64, activation='relu', input_shape=(784,)),\n",
        "    tf.keras.layers.Dense(emb_size, activation='sigmoid')\n",
        "])\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6gbK6iWO_Jwh",
        "outputId": "c27e940d-7d16-4b50-d56e-9de10f701633"
      },
      "source": [
        "embedding_model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense (Dense)                (None, 64)                50240     \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 64)                4160      \n",
            "=================================================================\n",
            "Total params: 54,400\n",
            "Trainable params: 54,400\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q-2s5xuG_8aX",
        "outputId": "b51586ea-258e-42ee-8b7b-6160e0dffe62"
      },
      "source": [
        "img= np.expand_dims(x_train[0], axis=0)\n",
        "example_emb = embedding_model.predict(img)[0]\n",
        "\n",
        "print(\"Embedding is:\",example_emb)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Embedding is: [0.5558166  0.4789469  0.5317682  0.5817168  0.5125536  0.52936715\n",
            " 0.44506085 0.41482466 0.4883641  0.5178479  0.48641148 0.46334848\n",
            " 0.56566983 0.50364685 0.58933526 0.5548806  0.5400239  0.5068117\n",
            " 0.50458384 0.5324708  0.5325529  0.5761824  0.44169062 0.55862546\n",
            " 0.45468074 0.6136361  0.39155254 0.55574286 0.58217895 0.45600674\n",
            " 0.57752913 0.6132797  0.42865747 0.38268888 0.49420983 0.38850552\n",
            " 0.42166543 0.4359734  0.5522154  0.54921347 0.523246   0.47620258\n",
            " 0.3711207  0.4876512  0.5719194  0.5678689  0.5033949  0.549955\n",
            " 0.4693066  0.4094522  0.44309568 0.49248514 0.5982806  0.50349045\n",
            " 0.34433186 0.48971647 0.50057507 0.63986444 0.52874315 0.5199904\n",
            " 0.5430845  0.45710063 0.56601626 0.52948743]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CntA023QABMw"
      },
      "source": [
        "Creation of the Siamese Network starts here"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GwtpRPGbACjW",
        "outputId": "88fc15d6-387c-4306-89b5-dab3dd1bac10"
      },
      "source": [
        "    input_anchor = tf.keras.layers.Input(shape=(784,))\n",
        "    input_positive = tf.keras.layers.Input(shape=(784,))\n",
        "    input_negative = tf.keras.layers.Input(shape=(784,))\n",
        "\n",
        "    embedding_anchor = embedding_model(input_anchor)\n",
        "    embedding_positive = embedding_model(input_positive)\n",
        "    embedding_negative = embedding_model(input_negative)\n",
        "\n",
        "    output = tf.keras.layers.concatenate([embedding_anchor, embedding_positive, embedding_negative], axis=1) #Concat alll embeddings \n",
        "\n",
        "    siamnet = tf.keras.models.Model([input_anchor, input_positive, input_negative], output) \n",
        "    siamnet.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_3\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_4 (InputLayer)            [(None, 784)]        0                                            \n",
            "__________________________________________________________________________________________________\n",
            "input_5 (InputLayer)            [(None, 784)]        0                                            \n",
            "__________________________________________________________________________________________________\n",
            "input_6 (InputLayer)            [(None, 784)]        0                                            \n",
            "__________________________________________________________________________________________________\n",
            "sequential (Sequential)         (None, 64)           54400       input_4[0][0]                    \n",
            "                                                                 input_5[0][0]                    \n",
            "                                                                 input_6[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 192)          0           sequential[3][0]                 \n",
            "                                                                 sequential[4][0]                 \n",
            "                                                                 sequential[5][0]                 \n",
            "==================================================================================================\n",
            "Total params: 54,400\n",
            "Trainable params: 54,400\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 475
        },
        "id": "eLJTzbLAOx8P",
        "outputId": "efc42b59-dc5d-4ebe-a7fa-1ffd169087ab"
      },
      "source": [
        "tf.keras.utils.plot_model(siamnet,show_shapes=True,show_layer_names=True,expand_nested=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l2E0U1MUAFWu"
      },
      "source": [
        "alpha = 0.2\n",
        "\n",
        "def triplet_loss(y_true, y_pred):\n",
        "    anchor, positive, negative = y_pred[:,:emb_size], y_pred[:,emb_size:2*emb_size], y_pred[:,2*emb_size:]\n",
        "    positive_dist = tf.reduce_mean(tf.square(anchor - positive), axis=1)\n",
        "    negative_dist = tf.reduce_mean(tf.square(anchor - negative), axis=1)\n",
        "    return tf.maximum(positive_dist - negative_dist + alpha, 0.)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vn0fAsJJAP7P"
      },
      "source": [
        "def data_generator(batch_size=256):\n",
        "    while True:\n",
        "        x = create_batch(batch_size)\n",
        "        y = np.zeros((batch_size, 3*emb_size))\n",
        "        yield x, y"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SH1jEtSEIYX-"
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "from sklearn.decomposition import PCA\n",
        "\n",
        "class PCAPlotter(tf.keras.callbacks.Callback):\n",
        "    \n",
        "    def __init__(self, plt, embedding_model, x_test, y_test):\n",
        "        super(PCAPlotter, self).__init__()\n",
        "        self.embedding_model = embedding_model\n",
        "        self.x_test = x_test\n",
        "        self.y_test = y_test\n",
        "        self.fig = plt.figure(figsize=(9, 4))\n",
        "        self.ax1 = plt.subplot(1, 2, 1)\n",
        "        self.ax2 = plt.subplot(1, 2, 2)\n",
        "        plt.ion()\n",
        "        \n",
        "        self.losses = []\n",
        "    \n",
        "    def plot(self, epoch=None, plot_loss=False):\n",
        "        x_test_embeddings = self.embedding_model.predict(self.x_test)\n",
        "        pca_out = PCA(n_components=2).fit_transform(x_test_embeddings)\n",
        "        self.ax1.clear()\n",
        "        self.ax1.scatter(pca_out[:, 0], pca_out[:, 1], c=self.y_test, cmap='seismic')\n",
        "        if plot_loss:\n",
        "            self.ax2.clear()\n",
        "            self.ax2.plot(range(epoch), self.losses)\n",
        "            self.ax2.set_xlabel('Epochs')\n",
        "            self.ax2.set_ylabel('Loss')\n",
        "        self.fig.canvas.draw()\n",
        "    \n",
        "    def on_train_begin(self, logs=None):\n",
        "        self.losses = []\n",
        "        self.fig.show()\n",
        "        self.fig.canvas.draw()\n",
        "        self.plot()\n",
        "        \n",
        "    def on_epoch_end(self, epoch, logs=None):\n",
        "        self.losses.append(logs.get('loss'))\n",
        "        self.plot(epoch+1, plot_loss=True)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "id": "c7eDA40MAStu",
        "outputId": "ffdcf9ba-0323-4824-b4c9-89d91122b08f"
      },
      "source": [
        "batch_size = 2048\n",
        "epochs = 10\n",
        "steps_per_epoch = int(x_train.shape[0]/batch_size)   # 60K/2048 =~30 steps per epoch\n",
        "\n",
        "siamnet.compile(loss=triplet_loss, optimizer='adam')\n",
        "\n",
        "_= siamnet.fit(\n",
        "    data_generator(batch_size),\n",
        "    steps_per_epoch=steps_per_epoch,\n",
        "    epochs=epochs, verbose=False,\n",
        "    callbacks=[\n",
        "        PCAPlotter(\n",
        "            plt, embedding_model,\n",
        "            x_test[:10000], y_test[:10000]\n",
        "        )]\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iT2z7A6wKtje"
      },
      "source": [
        "embedding_model.save('embedding_model.h5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C-e0G8OkQVHX",
        "outputId": "5bd88950-d859-49fd-eb04-ae1bbab6f289"
      },
      "source": [
        "siamnet.layers[3].layers[1].get_weights()[1].shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(64,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 97
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wqGx0vSGQpb_",
        "outputId": "54213ac1-b9e5-4449-e9e2-709eb4526739"
      },
      "source": [
        "siamnet.layers[3].get_weights()[3].shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(64,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "23pf-4y_RoUn"
      },
      "source": [
        "siamnet.save_weights('siamnet_weights.h5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9ZTLM_t3SsSt",
        "outputId": "62c849ce-6ad9-4f3a-d73d-6489c6ed1d0a"
      },
      "source": [
        "embedding_model.layers[1].get_weights()[1].shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(64,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 110
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aYfFBB_vSuLp"
      },
      "source": [
        "siamnet.load_weights('siamnet_weights.h5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_JlPABAdWouL",
        "outputId": "ecfd2918-82e6-4d5c-98a3-5f5f1e70d658"
      },
      "source": [
        "siamnet.layers[3].layers[1].get_weights()[1].shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(64,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 136
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cExQ3FkNVgD_"
      },
      "source": [
        "    input_1 = tf.keras.layers.Input(shape=(784,))\n",
        "    input_2 = tf.keras.layers.Input(shape=(784,))\n",
        "    "
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}