CM_VGG16_LSTM_Test.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CM_VGG16_LSTM_Test.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/TruongGiangVu/VGG16_RNN_LSTM_SignLanguageRecognition/blob/main/CM_VGG16_LSTM_Test.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vcOqjcCwbm0m"
      },
      "source": [
        "# Import"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oQibkISPqXeM",
        "outputId": "84dabc92-2bfa-43e2-8842-10327dd74090"
      },
      "source": [
        "!pip uninstall tensorflow\n",
        "!pip install --no-cache-dir tensorflow==2.2.0"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Uninstalling tensorflow-2.4.1:\n",
            "  Would remove:\n",
            "    /usr/local/bin/estimator_ckpt_converter\n",
            "    /usr/local/bin/import_pb_to_tensorboard\n",
            "    /usr/local/bin/saved_model_cli\n",
            "    /usr/local/bin/tensorboard\n",
            "    /usr/local/bin/tf_upgrade_v2\n",
            "    /usr/local/bin/tflite_convert\n",
            "    /usr/local/bin/toco\n",
            "    /usr/local/bin/toco_from_protos\n",
            "    /usr/local/lib/python3.7/dist-packages/tensorflow-2.4.1.dist-info/*\n",
            "    /usr/local/lib/python3.7/dist-packages/tensorflow/*\n",
            "Proceed (y/n)? y\n",
            "  Successfully uninstalled tensorflow-2.4.1\n",
            "Collecting tensorflow==2.2.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/4c/1a/0d79814736cfecc825ab8094b39648cc9c46af7af1bae839928acb73b4dd/tensorflow-2.2.0-cp37-cp37m-manylinux2010_x86_64.whl (516.2MB)\n",
            "\u001b[K     |████████████████████████████████| 516.2MB 1.5MB/s \n",
            "\u001b[?25hRequirement already satisfied: scipy==1.4.1; python_version >= \"3\" in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.4.1)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.1.0)\n",
            "Requirement already satisfied: astunparse==1.6.3 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.6.3)\n",
            "Requirement already satisfied: numpy<2.0,>=1.16.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.19.5)\n",
            "Requirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (0.12.0)\n",
            "Collecting tensorflow-estimator<2.3.0,>=2.2.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/a4/f5/926ae53d6a226ec0fda5208e0e581cffed895ccc89e36ba76a8e60895b78/tensorflow_estimator-2.2.0-py2.py3-none-any.whl (454kB)\n",
            "\u001b[K     |████████████████████████████████| 460kB 28.7MB/s \n",
            "\u001b[?25hRequirement already satisfied: protobuf>=3.8.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (3.12.4)\n",
            "Requirement already satisfied: wrapt>=1.11.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.12.1)\n",
            "Requirement already satisfied: gast==0.3.3 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (0.3.3)\n",
            "Requirement already satisfied: keras-preprocessing>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.1.2)\n",
            "Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.32.0)\n",
            "Requirement already satisfied: wheel>=0.26; python_version >= \"3\" in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (0.36.2)\n",
            "Collecting tensorboard<2.3.0,>=2.2.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/1d/74/0a6fcb206dcc72a6da9a62dd81784bfdbff5fedb099982861dc2219014fb/tensorboard-2.2.2-py3-none-any.whl (3.0MB)\n",
            "\u001b[K     |████████████████████████████████| 3.0MB 39.4MB/s \n",
            "\u001b[?25hRequirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (3.3.0)\n",
            "Requirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (1.15.0)\n",
            "Requirement already satisfied: h5py<2.11.0,>=2.10.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (2.10.0)\n",
            "Requirement already satisfied: google-pasta>=0.1.8 in /usr/local/lib/python3.7/dist-packages (from tensorflow==2.2.0) (0.2.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from protobuf>=3.8.0->tensorflow==2.2.0) (56.1.0)\n",
            "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (0.4.4)\n",
            "Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (1.8.0)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (2.23.0)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (1.0.1)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (3.3.4)\n",
            "Requirement already satisfied: google-auth<2,>=1.6.3 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (1.28.1)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.7/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (1.3.0)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (3.0.4)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (2.10)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (2020.12.5)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (1.24.3)\n",
            "Requirement already satisfied: importlib-metadata; python_version < \"3.8\" in /usr/local/lib/python3.7/dist-packages (from markdown>=2.6.8->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (3.10.1)\n",
            "Requirement already satisfied: rsa<5,>=3.1.4; python_version >= \"3.6\" in /usr/local/lib/python3.7/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (4.7.2)\n",
            "Requirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (4.2.1)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from google-auth<2,>=1.6.3->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (0.2.8)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (3.1.0)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata; python_version < \"3.8\"->markdown>=2.6.8->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (3.4.1)\n",
            "Requirement already satisfied: typing-extensions>=3.6.4; python_version < \"3.8\" in /usr/local/lib/python3.7/dist-packages (from importlib-metadata; python_version < \"3.8\"->markdown>=2.6.8->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (3.7.4.3)\n",
            "Requirement already satisfied: pyasn1>=0.1.3 in /usr/local/lib/python3.7/dist-packages (from rsa<5,>=3.1.4; python_version >= \"3.6\"->google-auth<2,>=1.6.3->tensorboard<2.3.0,>=2.2.0->tensorflow==2.2.0) (0.4.8)\n",
            "Installing collected packages: tensorflow-estimator, tensorboard, tensorflow\n",
            "  Found existing installation: tensorflow-estimator 2.4.0\n",
            "    Uninstalling tensorflow-estimator-2.4.0:\n",
            "      Successfully uninstalled tensorflow-estimator-2.4.0\n",
            "  Found existing installation: tensorboard 2.4.1\n",
            "    Uninstalling tensorboard-2.4.1:\n",
            "      Successfully uninstalled tensorboard-2.4.1\n",
            "Successfully installed tensorboard-2.2.2 tensorflow-2.2.0 tensorflow-estimator-2.2.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5fqXERKxbaaX",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7c07db56-4cb1-4e73-8c45-a0306cbf66b2"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8Ge2FCREbqke"
      },
      "source": [
        "# common library \n",
        "import numpy as np\n",
        "import os\n",
        "import time\n",
        "import cv2\n",
        "import matplotlib.pyplot as plt\n",
        "import glob\n",
        "from google.colab.patches import cv2_imshow\n",
        "\n",
        "# tensorflow library\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.models import Sequential, Model\n",
        "from keras.layers import Dense, GRU, LSTM, Flatten, TimeDistributed, Flatten, BatchNormalization, Activation, Dropout\n",
        "from keras.layers import Conv2D, BatchNormalization, MaxPool2D, GlobalMaxPool2D, MaxPooling2D\n",
        "# from keras.layers import TimeDistributed, GRU, Dense, Dropout\n",
        "from keras.layers.convolutional import Conv3D, MaxPooling3D\n",
        "from keras.callbacks import ModelCheckpoint, ReduceLROnPlateau, EarlyStopping\n",
        "from keras.losses import categorical_crossentropy\n",
        "from keras.optimizers import Adam\n",
        "\n",
        "from keras.applications.vgg16 import VGG16\n",
        "\n",
        "# reading data\n",
        "import csv\n",
        "import threading\n",
        "\n",
        "# Label encoder\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from keras.utils import np_utils"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9IQwb27bbvUl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b126c3b3-f143-42cc-bdc3-93b96ede5791"
      },
      "source": [
        "# change path\n",
        "import os\n",
        "os.chdir('/content/drive/MyDrive/VSLProject')\n",
        "print(os.getcwd())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/MyDrive/VSLProject\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ur7qEYnfbxdl"
      },
      "source": [
        "# Global params"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QAZ50D6pbz_u"
      },
      "source": [
        "# some global params\n",
        "SIZE = (64, 64) \n",
        "\n",
        "# SIZE = (224, 224) \n",
        "CHANNELS = 3\n",
        "NBFRAME = 10\n",
        "BS = 20\n",
        "\n",
        "NUM_THREADS = 8\n",
        "\n",
        "#preprocessing\n",
        "LENGTH_TRIM = 130\n",
        "classes_file = 'classes.npy'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3H20UoCVb30u"
      },
      "source": [
        "#Funtion"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w0ihc1ijcNT0"
      },
      "source": [
        "## Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GwEKz4I4b5Et"
      },
      "source": [
        "# resize image by dsize = (128, 128)\n",
        "def resizeImage(img, dsize):\n",
        "  sizeImg = (img.shape[0],img.shape[1])\n",
        "  if sizeImg != dsize and dsize != (None, None):\n",
        "    img = cv2.resize(img,dsize)\n",
        "  return img\n",
        "\n",
        "# remove left right background\n",
        "def removeTrimBackground(img, length = 120):\n",
        "  img = img[:,length:-1*length]\n",
        "  return img\n",
        "\n",
        "def resizeScale(img, scale = 30):\n",
        "  scale_percent = scale # percent of original size\n",
        "  width = int(img.shape[1] * scale_percent / 100)\n",
        "  height = int(img.shape[0] * scale_percent / 100)\n",
        "  dim = (width, height)\n",
        "    \n",
        "  # resize image\n",
        "  resized = cv2.resize(img, dim, interpolation = cv2.INTER_AREA)\n",
        "  return resized\n",
        "  \n",
        "def preprocessingImg(img, dsize = (128,128)):\n",
        "  # img = human_detection(img)\n",
        "  img = removeTrimBackground(img, LENGTH_TRIM)\n",
        "  img = resizeImage(img, dsize)\n",
        "  return img"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q6dxT7dWcPzN"
      },
      "source": [
        "## Video to frame"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-42njMG7cMok"
      },
      "source": [
        "class MyVideo:\n",
        "  def __init__(self, path='', root_path=''):\n",
        "    self.path = path\n",
        "    self.root = root_path\n",
        "    self.video = cv2.VideoCapture(path)\n",
        "    self.name = path.split(\"/\")[-1]\n",
        "    self.label = path.split(\"/\")[-2]\n",
        "  def getVideo(self):\n",
        "    return self.video\n",
        "  def getFullPath(self):\n",
        "    return self.root + '/' + self.path\n",
        "  def getFileName(self):\n",
        "    return self.name\n",
        "  def getFrameCount(self):\n",
        "    return int(self.video.get(cv2.CAP_PROP_FRAME_COUNT))\n",
        "  def getLabel(self):\n",
        "    return self.label\n",
        "  def release(self):\n",
        "    self.video.release()\n",
        "    del self\n",
        "\n",
        "\n",
        "def videotoframe(myVideo, dsize=(128,128) , numframe=5, start=0, stop=0):\n",
        "  listframe = []\n",
        "  length = myVideo.getFrameCount()\n",
        "\n",
        "  currentFrame = -1\n",
        "  skipframe = int((length - start - stop)/numframe)\n",
        "  takeframe = skipframe + start\n",
        "  taken = 0\n",
        "  # print('video',video,'length', length, 'skip', skipframe)\n",
        "  video = myVideo.getVideo()\n",
        "  while(True):\n",
        "      # Capture frame-by-frame\n",
        "      ret, frame = video.read()\n",
        "      currentFrame += 1\n",
        "\n",
        "      if ret == False:\n",
        "          break\n",
        "      if currentFrame < takeframe:\n",
        "          continue\n",
        "      if taken >= numframe:\n",
        "          break\n",
        "      frame = resizeScale(frame)\n",
        "      frame = removeTrimBackground(frame, LENGTH_TRIM)\n",
        "      frame = resizeImage(frame, dsize)\n",
        "      listframe.append(frame)\n",
        "\n",
        "      takeframe += skipframe\n",
        "      taken += 1\n",
        "  # When everything done, release the capture\n",
        "  myVideo.release()\n",
        "  cv2.destroyAllWindows()\n",
        "  return listframe\n",
        "\n",
        "def preprocessingVideo(myVideo):\n",
        "  frames = videotoframe(myVideo, SIZE, NBFRAME, 10, 30)\n",
        "  frames = np.array(frames)\n",
        "  frames = frames / 255.\n",
        "  return frames\n",
        "\n",
        "def predictOneVideo(video, classes):\n",
        "  # preprocessing video path\n",
        "  if type(video) is not MyVideo:\n",
        "    video = MyVideo(video)\n",
        "  item = preprocessingVideo(video)\n",
        "  item = item[None,:]  # [640,10,64,64,3]\n",
        "  try:\n",
        "    # a = model.predict_classes(item) # old version\n",
        "    predicted = np.argmax(model.predict(item), axis=-1) # [10] [15,12]\n",
        "    return classes[predicted[0]]  # [16] accept\n",
        "  except: \n",
        "    print('Video', myVideo.name,'has error')\n",
        "    return 'Null'\n",
        "  # print(predicted)\n",
        "\n",
        "def preparePath(path, csv_file = ''):\n",
        "  paths = []\n",
        "  if csv_file == '':\n",
        "    videos = os.listdir(path)\n",
        "    for video in videos:\n",
        "        video_path = os.path.join(path, video)\n",
        "        paths.append(video_path)\n",
        "  else:\n",
        "    with open(csv_file) as csvfile: # read path from csv\n",
        "      reader = csv.reader(csvfile)\n",
        "      for row in reader:\n",
        "        video_path = os.path.join(path, row[0])\n",
        "        paths.append(video_path)\n",
        "\n",
        "  print(len(paths))\n",
        "  return paths"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9Jf9osZQcW0k"
      },
      "source": [
        "## Predict video"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xfAACwH2cVqs"
      },
      "source": [
        "def predictVideoOneTime(paths, classes, verbose = 0):\n",
        "  data = []\n",
        "  names = []\n",
        "  expecteds = []\n",
        "  n = len(videos)\n",
        "  for path in paths:\n",
        "    myVideo = MyVideo(path)\n",
        "    item = preprocessingVideo(myVideo)\n",
        "    data.append(item)\n",
        "    names.append(myVideo.getFileName())\n",
        "    expecteds.append(myVideo.getLabel())\n",
        "\n",
        "  data = np.array(data)\n",
        "  # print(data.shape)\n",
        "  predicted = np.argmax(model.predict(data), axis=-1)\n",
        "  # print(predicted)\n",
        "  # print(classes[predicted])\n",
        "  dic = {}\n",
        "  count = 0\n",
        "  for i in range(n):\n",
        "    expected = expecteds[i]\n",
        "    text = 'Video: '+ names[i] + ' Expected: '+ expected + ' Predicted: '+ classes[predicted[i]]\n",
        "    if expected not in dic:\n",
        "      dic[expected] = 0\n",
        "    if expected == classes[predicted[i]]:\n",
        "      count = count + 1\n",
        "      dic[expected] = dic[expected] + 1\n",
        "    else:\n",
        "      text = '\\x1b[31m'+ text + '\\x1b[0m'\n",
        "    \n",
        "    if verbose == 1 : \n",
        "      print(text)\n",
        "  print()\n",
        "  print('Accuray:',str(count)+'/'+str(n), ',',str(count/n))\n",
        "  return dic\n",
        "\n",
        "def predictVideoOneByOne(paths, classes, verbose = 0):\n",
        "  dic = {}\n",
        "\n",
        "  count = 0\n",
        "  n = len(paths)\n",
        "  index = 1\n",
        "  for video in paths:\n",
        "    myVideo = MyVideo(video)\n",
        "    predicted = predictOneVideo( myVideo, classes)\n",
        "    expected = myVideo.getLabel()\n",
        "    print(index, end=' ')\n",
        "    text = 'Video: '+ myVideo.getFileName() + ' Expected: '+  expected + ' Predicted: '+ predicted\n",
        "    if expected not in dic:\n",
        "      dic[expected] = 0\n",
        "    if expected == predicted:\n",
        "      count = count + 1\n",
        "      dic[expected] = dic[expected] + 1\n",
        "    else:\n",
        "      text = '\\x1b[31m'+ text + '\\x1b[0m'\n",
        "    index = index + 1\n",
        "    \n",
        "    if verbose == 1 : \n",
        "      print(text)\n",
        "    \n",
        "  print()\n",
        "  print('Accuray:',str(count)+'/'+str(n), ',',str(count/n))\n",
        "  return dic\n",
        "\n",
        "def predictVideoOneByOneReturn2dDict(paths, classes, verbose = 0):\n",
        "  dic = {}    # to make a confusion maxtrix, im use a 2D dictionary\n",
        "  count = 0\n",
        "  n = len(paths)\n",
        "  index = 0\n",
        "  for video in paths:\n",
        "    myVideo = MyVideo(video)\n",
        "    predicted = predictOneVideo( myVideo, classes)\n",
        "    expected = myVideo.getLabel()\n",
        "    print(index+1, end=' ')\n",
        "    text = 'Video: '+ myVideo.getFileName() + ' Expected: '+  expected + ' Predicted: '+ predicted\n",
        "    if expected not in dic:\n",
        "      dic[expected] = dict()\n",
        "    if predicted not in dic[expected]:\n",
        "      dic[expected][predicted] = 0\n",
        "    if expected == predicted:\n",
        "      count = count + 1\n",
        "      # dic[expected] = dic[expected] + 1\n",
        "    else:\n",
        "      text = '\\x1b[31m'+ text + '\\x1b[0m'\n",
        "    \n",
        "    # process onfusion matrix array\n",
        "    dic[expected][predicted] += 1\n",
        "\n",
        "\n",
        "    index = index + 1\n",
        "    if verbose == 1 : \n",
        "      print(text)\n",
        "    \n",
        "  print()\n",
        "  print('Accuray:',str(count)+'/'+str(n), ',',str(count/n))\n",
        "  return dic\n",
        "\n",
        "def predictVideo(path, csv_file, classes, verbose = 0 ,isOneTime = False):\n",
        "  paths = preparePath(path, csv_file)\n",
        "  \n",
        "  if isOneTime == True:\n",
        "    dic = predictVideoOneTime(paths, classes, verbose)\n",
        "  else:\n",
        "    dic = predictVideoOneByOneReturn2dDict(paths, classes, verbose)\n",
        "  return dic"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JkH-BQ4rcEj0"
      },
      "source": [
        "# Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "osyGIM78b85p"
      },
      "source": [
        "def load_VGG16_model(shape=(112, 112, 3)):\n",
        "  base_model = VGG16(weights='imagenet', include_top=False, input_shape=shape)\n",
        "  print(\"Model loaded..!\")\n",
        "  base_model.summary()\n",
        "\n",
        "  # model.add(Flatten())\n",
        "  output1 = GlobalMaxPool2D()\n",
        "  output2 = Flatten()\n",
        "  module = keras.Sequential([base_model, output1, output2])\n",
        "  module.summary()\n",
        "  return module\n",
        "\n",
        "def LSTMModel(shape=(5, 112, 112, 3), nbout=3):\n",
        "  model = Sequential()\n",
        "  convnet = load_VGG16_model(shape[1:])\n",
        "  # convnet = VGG16_small(shape[1:])\n",
        "  model.add(TimeDistributed(convnet, input_shape=shape))\n",
        "\n",
        "  model.add(LSTM(256, dropout=0.2))\n",
        "  model.add(Dense(1024, activation='relu'))\n",
        "  model.add(Dropout(0.5))\n",
        "  model.add(Dense(nbout, activation='softmax'))\n",
        "  sgd = keras.optimizers.SGD(lr=0.00005, decay = 1e-6, momentum=0.9, nesterov=True)\n",
        "  # sgd = keras.optimizers.SGD( momentum=0.9, nesterov=True)\n",
        "  model.compile(optimizer=sgd, loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "  return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cbBbkrelcaZF"
      },
      "source": [
        "# Main"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SgqpS_RZcb_F",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "99eda42e-6e32-479d-814d-92bb9a8b0e2f"
      },
      "source": [
        "# read classes from file\n",
        "classes_file = 'classes.npy'\n",
        "encoder = LabelEncoder()\n",
        "encoder.classes_ = np.load(classes_file)\n",
        "classes = np.array(encoder.classes_)\n",
        "num_classes = len(classes)\n",
        "INSHAPE=(NBFRAME,) + SIZE + (CHANNELS,) # (2560, 5, 112, 112, 3)\n",
        "print(classes)\n",
        "# print(INSHAPE) \n",
        "model = LSTMModel(INSHAPE, num_classes)\n",
        "model.load_weights('chkp/weights-LSTM-095625-08_05.hdf5')\n",
        "\n",
        "print(num_classes,classes)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['Accept' 'Appear' 'Argentina' 'Away' 'Barbecue' 'Bathe' 'Birthday'\n",
            " 'Bitter' 'Born' 'Breakfast' 'Bright' 'Buy' 'Call' 'Candy' 'Catch'\n",
            " 'Chewing-gum' 'Coin' 'Colors' 'Copy' 'Country' 'Dance' 'Deaf' 'Drawer'\n",
            " 'Enemy' 'Find' 'Food' 'Give' 'Green' 'Help' 'Hungry' 'Last name' 'Learn'\n",
            " 'Light-blue' 'Man' 'Map' 'Milk' 'Mock' 'Music' 'Name' 'None' 'Opaque'\n",
            " 'Patience' 'Perfume' 'Photo' 'Realize' 'Red' 'Red_2' 'Rice' 'Run' 'Ship'\n",
            " 'Shut down' 'Skimmer' 'Son' 'Spaghetti' 'Sweet milk' 'Thanks' 'To land'\n",
            " 'Trap' 'Uruguay' 'Water' 'Where' 'Women' 'Yellow' 'Yogurt']\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "58892288/58889256 [==============================] - 1s 0us/step\n",
            "Model loaded..!\n",
            "Model: \"vgg16\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         [(None, 64, 64, 3)]       0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 64, 64, 64)        1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 64, 64, 64)        36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 32, 32, 64)        0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 32, 32, 128)       73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 32, 32, 128)       147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 16, 16, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 16, 16, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 16, 16, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 16, 16, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 8, 8, 256)         0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 8, 8, 512)         1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 8, 8, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 8, 8, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 4, 4, 512)         0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 2, 2, 512)         0         \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 14,714,688\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "vgg16 (Model)                (None, 2, 2, 512)         14714688  \n",
            "_________________________________________________________________\n",
            "global_max_pooling2d (Global (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 512)               0         \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 14,714,688\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "64 ['Accept' 'Appear' 'Argentina' 'Away' 'Barbecue' 'Bathe' 'Birthday'\n",
            " 'Bitter' 'Born' 'Breakfast' 'Bright' 'Buy' 'Call' 'Candy' 'Catch'\n",
            " 'Chewing-gum' 'Coin' 'Colors' 'Copy' 'Country' 'Dance' 'Deaf' 'Drawer'\n",
            " 'Enemy' 'Find' 'Food' 'Give' 'Green' 'Help' 'Hungry' 'Last name' 'Learn'\n",
            " 'Light-blue' 'Man' 'Map' 'Milk' 'Mock' 'Music' 'Name' 'None' 'Opaque'\n",
            " 'Patience' 'Perfume' 'Photo' 'Realize' 'Red' 'Red_2' 'Rice' 'Run' 'Ship'\n",
            " 'Shut down' 'Skimmer' 'Son' 'Spaghetti' 'Sweet milk' 'Thanks' 'To land'\n",
            " 'Trap' 'Uruguay' 'Water' 'Where' 'Women' 'Yellow' 'Yogurt']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YhCe1THYcfG1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "77bb5d0d-1be1-4127-ad2c-8722f5c83156"
      },
      "source": [
        "start = time.time()\n",
        "\n",
        "res = predictVideo(\n",
        "    path='Datasets/lsa64_split/test', \n",
        "    csv_file='Datasets/lsa64_videoperfolder/test.csv', \n",
        "    classes=classes,\n",
        "    verbose=1,\n",
        "    isOneTime = False)\n",
        "\n",
        "end = time.time()\n",
        "print('time:', end - start)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "640\n",
            "1 Video: 049_001_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "2 Video: 049_002_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "3 Video: 049_003_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "4 Video: 049_004_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "5 Video: 049_005_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "6 Video: 049_006_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "7 Video: 049_007_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "8 Video: 049_008_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "9 Video: 049_009_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "10 Video: 049_010_005.mp4 Expected: Yogurt Predicted: Yogurt\n",
            "11 Video: 009_001_005.mp4 Expected: Women Predicted: Women\n",
            "12 Video: 009_002_005.mp4 Expected: Women Predicted: Women\n",
            "13 Video: 009_003_005.mp4 Expected: Women Predicted: Women\n",
            "14 Video: 009_004_005.mp4 Expected: Women Predicted: Women\n",
            "15 Video: 009_005_005.mp4 Expected: Women Predicted: Women\n",
            "16 Video: 009_006_005.mp4 Expected: Women Predicted: Women\n",
            "17 Video: 009_007_005.mp4 Expected: Women Predicted: Women\n",
            "18 Video: 009_008_005.mp4 Expected: Women Predicted: Women\n",
            "19 \u001b[31mVideo: 009_009_005.mp4 Expected: Women Predicted: Drawer\u001b[0m\n",
            "20 Video: 009_010_005.mp4 Expected: Women Predicted: Women\n",
            "21 Video: 028_001_005.mp4 Expected: Where Predicted: Where\n",
            "22 Video: 028_002_005.mp4 Expected: Where Predicted: Where\n",
            "23 Video: 028_003_005.mp4 Expected: Where Predicted: Where\n",
            "24 Video: 028_004_005.mp4 Expected: Where Predicted: Where\n",
            "25 Video: 028_005_005.mp4 Expected: Where Predicted: Where\n",
            "26 Video: 028_006_005.mp4 Expected: Where Predicted: Where\n",
            "27 Video: 028_007_005.mp4 Expected: Where Predicted: Where\n",
            "28 Video: 028_008_005.mp4 Expected: Where Predicted: Where\n",
            "29 Video: 028_009_005.mp4 Expected: Where Predicted: Where\n",
            "30 Video: 028_010_005.mp4 Expected: Where Predicted: Where\n",
            "31 Video: 043_001_005.mp4 Expected: Trap Predicted: Trap\n",
            "32 Video: 043_002_005.mp4 Expected: Trap Predicted: Trap\n",
            "33 Video: 043_003_005.mp4 Expected: Trap Predicted: Trap\n",
            "34 Video: 043_004_005.mp4 Expected: Trap Predicted: Trap\n",
            "35 Video: 043_005_005.mp4 Expected: Trap Predicted: Trap\n",
            "36 Video: 043_006_005.mp4 Expected: Trap Predicted: Trap\n",
            "37 Video: 043_007_005.mp4 Expected: Trap Predicted: Trap\n",
            "38 Video: 043_008_005.mp4 Expected: Trap Predicted: Trap\n",
            "39 Video: 043_009_005.mp4 Expected: Trap Predicted: Trap\n",
            "40 Video: 043_010_005.mp4 Expected: Trap Predicted: Trap\n",
            "41 Video: 025_001_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "42 Video: 025_002_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "43 Video: 025_003_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "44 Video: 025_004_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "45 Video: 025_005_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "46 Video: 025_006_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "47 Video: 025_007_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "48 Video: 025_008_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "49 Video: 025_009_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "50 Video: 025_010_005.mp4 Expected: Uruguay Predicted: Uruguay\n",
            "51 Video: 004_001_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "52 Video: 004_002_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "53 Video: 004_003_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "54 Video: 004_004_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "55 Video: 004_005_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "56 Video: 004_006_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "57 Video: 004_007_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "58 Video: 004_008_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "59 Video: 004_009_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "60 Video: 004_010_005.mp4 Expected: Yellow Predicted: Yellow\n",
            "61 Video: 022_001_005.mp4 Expected: Water Predicted: Water\n",
            "62 Video: 022_002_005.mp4 Expected: Water Predicted: Water\n",
            "63 Video: 022_003_005.mp4 Expected: Water Predicted: Water\n",
            "64 Video: 022_004_005.mp4 Expected: Water Predicted: Water\n",
            "65 Video: 022_005_005.mp4 Expected: Water Predicted: Water\n",
            "66 Video: 022_006_005.mp4 Expected: Water Predicted: Water\n",
            "67 Video: 022_007_005.mp4 Expected: Water Predicted: Water\n",
            "68 Video: 022_008_005.mp4 Expected: Water Predicted: Water\n",
            "69 Video: 022_009_005.mp4 Expected: Water Predicted: Water\n",
            "70 Video: 022_010_005.mp4 Expected: Water Predicted: Water\n",
            "71 Video: 051_001_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "72 Video: 051_002_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "73 Video: 051_003_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "74 Video: 051_004_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "75 Video: 051_005_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "76 Video: 051_006_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "77 Video: 051_007_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "78 Video: 051_008_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "79 Video: 051_009_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "80 Video: 051_010_005.mp4 Expected: Thanks Predicted: Thanks\n",
            "81 \u001b[31mVideo: 018_001_005.mp4 Expected: Skimmer Predicted: Bright\u001b[0m\n",
            "82 \u001b[31mVideo: 018_002_005.mp4 Expected: Skimmer Predicted: Milk\u001b[0m\n",
            "83 Video: 018_003_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "84 Video: 018_004_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "85 Video: 018_005_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "86 \u001b[31mVideo: 018_006_005.mp4 Expected: Skimmer Predicted: Bright\u001b[0m\n",
            "87 Video: 018_007_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "88 Video: 018_008_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "89 Video: 018_009_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "90 Video: 018_010_005.mp4 Expected: Skimmer Predicted: Skimmer\n",
            "91 Video: 037_001_005.mp4 Expected: Ship Predicted: Ship\n",
            "92 Video: 037_002_005.mp4 Expected: Ship Predicted: Ship\n",
            "93 Video: 037_003_005.mp4 Expected: Ship Predicted: Ship\n",
            "94 Video: 037_004_005.mp4 Expected: Ship Predicted: Ship\n",
            "95 Video: 037_005_005.mp4 Expected: Ship Predicted: Ship\n",
            "96 Video: 037_006_005.mp4 Expected: Ship Predicted: Ship\n",
            "97 Video: 037_007_005.mp4 Expected: Ship Predicted: Ship\n",
            "98 Video: 037_008_005.mp4 Expected: Ship Predicted: Ship\n",
            "99 Video: 037_009_005.mp4 Expected: Ship Predicted: Ship\n",
            "100 Video: 037_010_005.mp4 Expected: Ship Predicted: Ship\n",
            "101 Video: 054_001_005.mp4 Expected: To land Predicted: To land\n",
            "102 Video: 054_002_005.mp4 Expected: To land Predicted: To land\n",
            "103 Video: 054_003_005.mp4 Expected: To land Predicted: To land\n",
            "104 Video: 054_004_005.mp4 Expected: To land Predicted: To land\n",
            "105 Video: 054_005_005.mp4 Expected: To land Predicted: To land\n",
            "106 Video: 054_006_005.mp4 Expected: To land Predicted: To land\n",
            "107 Video: 054_007_005.mp4 Expected: To land Predicted: To land\n",
            "108 Video: 054_008_005.mp4 Expected: To land Predicted: To land\n",
            "109 Video: 054_009_005.mp4 Expected: To land Predicted: To land\n",
            "110 Video: 054_010_005.mp4 Expected: To land Predicted: To land\n",
            "111 Video: 048_001_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "112 Video: 048_002_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "113 Video: 048_003_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "114 Video: 048_004_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "115 Video: 048_005_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "116 Video: 048_006_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "117 Video: 048_007_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "118 Video: 048_008_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "119 Video: 048_009_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "120 Video: 048_010_005.mp4 Expected: Spaghetti Predicted: Spaghetti\n",
            "121 Video: 011_001_005.mp4 Expected: Son Predicted: Son\n",
            "122 Video: 011_002_005.mp4 Expected: Son Predicted: Son\n",
            "123 Video: 011_003_005.mp4 Expected: Son Predicted: Son\n",
            "124 Video: 011_004_005.mp4 Expected: Son Predicted: Son\n",
            "125 Video: 011_005_005.mp4 Expected: Son Predicted: Son\n",
            "126 Video: 011_006_005.mp4 Expected: Son Predicted: Son\n",
            "127 Video: 011_007_005.mp4 Expected: Son Predicted: Son\n",
            "128 \u001b[31mVideo: 011_008_005.mp4 Expected: Son Predicted: Light-blue\u001b[0m\n",
            "129 Video: 011_009_005.mp4 Expected: Son Predicted: Son\n",
            "130 Video: 011_010_005.mp4 Expected: Son Predicted: Son\n",
            "131 Video: 044_001_005.mp4 Expected: Rice Predicted: Rice\n",
            "132 Video: 044_002_005.mp4 Expected: Rice Predicted: Rice\n",
            "133 Video: 044_003_005.mp4 Expected: Rice Predicted: Rice\n",
            "134 Video: 044_004_005.mp4 Expected: Rice Predicted: Rice\n",
            "135 Video: 044_005_005.mp4 Expected: Rice Predicted: Rice\n",
            "136 Video: 044_006_005.mp4 Expected: Rice Predicted: Rice\n",
            "137 Video: 044_007_005.mp4 Expected: Rice Predicted: Rice\n",
            "138 Video: 044_008_005.mp4 Expected: Rice Predicted: Rice\n",
            "139 Video: 044_009_005.mp4 Expected: Rice Predicted: Rice\n",
            "140 Video: 044_010_005.mp4 Expected: Rice Predicted: Rice\n",
            "141 Video: 061_001_005.mp4 Expected: Run Predicted: Run\n",
            "142 Video: 061_002_005.mp4 Expected: Run Predicted: Run\n",
            "143 Video: 061_003_005.mp4 Expected: Run Predicted: Run\n",
            "144 Video: 061_004_005.mp4 Expected: Run Predicted: Run\n",
            "145 Video: 061_005_005.mp4 Expected: Run Predicted: Run\n",
            "146 Video: 061_006_005.mp4 Expected: Run Predicted: Run\n",
            "147 Video: 061_007_005.mp4 Expected: Run Predicted: Run\n",
            "148 Video: 061_008_005.mp4 Expected: Run Predicted: Run\n",
            "149 Video: 061_009_005.mp4 Expected: Run Predicted: Run\n",
            "150 Video: 061_010_005.mp4 Expected: Run Predicted: Run\n",
            "151 Video: 020_001_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "152 Video: 020_002_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "153 Video: 020_003_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "154 Video: 020_004_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "155 Video: 020_005_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "156 Video: 020_006_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "157 Video: 020_007_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "158 Video: 020_008_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "159 Video: 020_009_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "160 Video: 020_010_005.mp4 Expected: Sweet milk Predicted: Sweet milk\n",
            "161 Video: 052_001_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "162 Video: 052_002_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "163 Video: 052_003_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "164 \u001b[31mVideo: 052_004_005.mp4 Expected: Shut down Predicted: Uruguay\u001b[0m\n",
            "165 Video: 052_005_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "166 \u001b[31mVideo: 052_006_005.mp4 Expected: Shut down Predicted: Deaf\u001b[0m\n",
            "167 Video: 052_007_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "168 Video: 052_008_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "169 Video: 052_009_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "170 Video: 052_010_005.mp4 Expected: Shut down Predicted: Shut down\n",
            "171 Video: 036_001_005.mp4 Expected: Music Predicted: Music\n",
            "172 Video: 036_002_005.mp4 Expected: Music Predicted: Music\n",
            "173 Video: 036_003_005.mp4 Expected: Music Predicted: Music\n",
            "174 Video: 036_004_005.mp4 Expected: Music Predicted: Music\n",
            "175 Video: 036_005_005.mp4 Expected: Music Predicted: Music\n",
            "176 Video: 036_006_005.mp4 Expected: Music Predicted: Music\n",
            "177 Video: 036_007_005.mp4 Expected: Music Predicted: Music\n",
            "178 Video: 036_008_005.mp4 Expected: Music Predicted: Music\n",
            "179 Video: 036_009_005.mp4 Expected: Music Predicted: Music\n",
            "180 Video: 036_010_005.mp4 Expected: Music Predicted: Music\n",
            "181 Video: 041_001_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "182 Video: 041_002_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "183 Video: 041_003_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "184 Video: 041_004_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "185 Video: 041_005_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "186 Video: 041_006_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "187 Video: 041_007_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "188 Video: 041_008_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "189 Video: 041_009_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "190 Video: 041_010_005.mp4 Expected: Perfume Predicted: Perfume\n",
            "191 Video: 029_001_005.mp4 Expected: Mock Predicted: Mock\n",
            "192 Video: 029_002_005.mp4 Expected: Mock Predicted: Mock\n",
            "193 Video: 029_003_005.mp4 Expected: Mock Predicted: Mock\n",
            "194 Video: 029_004_005.mp4 Expected: Mock Predicted: Mock\n",
            "195 Video: 029_005_005.mp4 Expected: Mock Predicted: Mock\n",
            "196 Video: 029_006_005.mp4 Expected: Mock Predicted: Mock\n",
            "197 Video: 029_007_005.mp4 Expected: Mock Predicted: Mock\n",
            "198 Video: 029_008_005.mp4 Expected: Mock Predicted: Mock\n",
            "199 Video: 029_009_005.mp4 Expected: Mock Predicted: Mock\n",
            "200 Video: 029_010_005.mp4 Expected: Mock Predicted: Mock\n",
            "201 Video: 039_001_005.mp4 Expected: Name Predicted: Name\n",
            "202 Video: 039_002_005.mp4 Expected: Name Predicted: Name\n",
            "203 Video: 039_003_005.mp4 Expected: Name Predicted: Name\n",
            "204 Video: 039_004_005.mp4 Expected: Name Predicted: Name\n",
            "205 Video: 039_005_005.mp4 Expected: Name Predicted: Name\n",
            "206 Video: 039_006_005.mp4 Expected: Name Predicted: Name\n",
            "207 Video: 039_007_005.mp4 Expected: Name Predicted: Name\n",
            "208 Video: 039_008_005.mp4 Expected: Name Predicted: Name\n",
            "209 Video: 039_009_005.mp4 Expected: Name Predicted: Name\n",
            "210 Video: 039_010_005.mp4 Expected: Name Predicted: Name\n",
            "211 Video: 032_001_005.mp4 Expected: Photo Predicted: Photo\n",
            "212 Video: 032_002_005.mp4 Expected: Photo Predicted: Photo\n",
            "213 Video: 032_003_005.mp4 Expected: Photo Predicted: Photo\n",
            "214 Video: 032_004_005.mp4 Expected: Photo Predicted: Photo\n",
            "215 Video: 032_005_005.mp4 Expected: Photo Predicted: Photo\n",
            "216 Video: 032_006_005.mp4 Expected: Photo Predicted: Photo\n",
            "217 Video: 032_007_005.mp4 Expected: Photo Predicted: Photo\n",
            "218 Video: 032_008_005.mp4 Expected: Photo Predicted: Photo\n",
            "219 Video: 032_009_005.mp4 Expected: Photo Predicted: Photo\n",
            "220 Video: 032_010_005.mp4 Expected: Photo Predicted: Photo\n",
            "221 Video: 040_001_005.mp4 Expected: Patience Predicted: Patience\n",
            "222 Video: 040_002_005.mp4 Expected: Patience Predicted: Patience\n",
            "223 Video: 040_003_005.mp4 Expected: Patience Predicted: Patience\n",
            "224 Video: 040_004_005.mp4 Expected: Patience Predicted: Patience\n",
            "225 Video: 040_005_005.mp4 Expected: Patience Predicted: Patience\n",
            "226 Video: 040_006_005.mp4 Expected: Patience Predicted: Patience\n",
            "227 Video: 040_007_005.mp4 Expected: Patience Predicted: Patience\n",
            "228 Video: 040_008_005.mp4 Expected: Patience Predicted: Patience\n",
            "229 Video: 040_009_005.mp4 Expected: Patience Predicted: Patience\n",
            "230 Video: 040_010_005.mp4 Expected: Patience Predicted: Patience\n",
            "231 Video: 001_001_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "232 Video: 001_002_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "233 Video: 001_003_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "234 Video: 001_004_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "235 \u001b[31mVideo: 001_005_005.mp4 Expected: Opaque Predicted: Bright\u001b[0m\n",
            "236 \u001b[31mVideo: 001_006_005.mp4 Expected: Opaque Predicted: Away\u001b[0m\n",
            "237 Video: 001_007_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "238 Video: 001_008_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "239 Video: 001_009_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "240 Video: 001_010_005.mp4 Expected: Opaque Predicted: Opaque\n",
            "241 Video: 038_001_005.mp4 Expected: None Predicted: None\n",
            "242 Video: 038_002_005.mp4 Expected: None Predicted: None\n",
            "243 Video: 038_003_005.mp4 Expected: None Predicted: None\n",
            "244 Video: 038_004_005.mp4 Expected: None Predicted: None\n",
            "245 Video: 038_005_005.mp4 Expected: None Predicted: None\n",
            "246 Video: 038_006_005.mp4 Expected: None Predicted: None\n",
            "247 Video: 038_007_005.mp4 Expected: None Predicted: None\n",
            "248 Video: 038_008_005.mp4 Expected: None Predicted: None\n",
            "249 Video: 038_009_005.mp4 Expected: None Predicted: None\n",
            "250 Video: 038_010_005.mp4 Expected: None Predicted: None\n",
            "251 Video: 002_001_005.mp4 Expected: Red Predicted: Red\n",
            "252 Video: 002_002_005.mp4 Expected: Red Predicted: Red\n",
            "253 Video: 002_003_005.mp4 Expected: Red Predicted: Red\n",
            "254 \u001b[31mVideo: 002_004_005.mp4 Expected: Red Predicted: Green\u001b[0m\n",
            "255 Video: 002_005_005.mp4 Expected: Red Predicted: Red\n",
            "256 Video: 002_006_005.mp4 Expected: Red Predicted: Red\n",
            "257 Video: 002_007_005.mp4 Expected: Red Predicted: Red\n",
            "258 Video: 002_008_005.mp4 Expected: Red Predicted: Red\n",
            "259 Video: 002_009_005.mp4 Expected: Red Predicted: Red\n",
            "260 \u001b[31mVideo: 002_010_005.mp4 Expected: Red Predicted: Green\u001b[0m\n",
            "261 Video: 062_001_005.mp4 Expected: Realize Predicted: Realize\n",
            "262 Video: 062_002_005.mp4 Expected: Realize Predicted: Realize\n",
            "263 Video: 062_003_005.mp4 Expected: Realize Predicted: Realize\n",
            "264 Video: 062_004_005.mp4 Expected: Realize Predicted: Realize\n",
            "265 Video: 062_005_005.mp4 Expected: Realize Predicted: Realize\n",
            "266 Video: 062_006_005.mp4 Expected: Realize Predicted: Realize\n",
            "267 Video: 062_007_005.mp4 Expected: Realize Predicted: Realize\n",
            "268 Video: 062_008_005.mp4 Expected: Realize Predicted: Realize\n",
            "269 Video: 062_009_005.mp4 Expected: Realize Predicted: Realize\n",
            "270 Video: 062_010_005.mp4 Expected: Realize Predicted: Realize\n",
            "271 Video: 003_001_005.mp4 Expected: Green Predicted: Green\n",
            "272 Video: 003_002_005.mp4 Expected: Green Predicted: Green\n",
            "273 Video: 003_003_005.mp4 Expected: Green Predicted: Green\n",
            "274 Video: 003_004_005.mp4 Expected: Green Predicted: Green\n",
            "275 \u001b[31mVideo: 003_005_005.mp4 Expected: Green Predicted: Red\u001b[0m\n",
            "276 Video: 003_006_005.mp4 Expected: Green Predicted: Green\n",
            "277 Video: 003_007_005.mp4 Expected: Green Predicted: Green\n",
            "278 Video: 003_008_005.mp4 Expected: Green Predicted: Green\n",
            "279 Video: 003_009_005.mp4 Expected: Green Predicted: Green\n",
            "280 \u001b[31mVideo: 003_010_005.mp4 Expected: Green Predicted: Red\u001b[0m\n",
            "281 Video: 012_001_005.mp4 Expected: Man Predicted: Man\n",
            "282 Video: 012_002_005.mp4 Expected: Man Predicted: Man\n",
            "283 Video: 012_003_005.mp4 Expected: Man Predicted: Man\n",
            "284 Video: 012_004_005.mp4 Expected: Man Predicted: Man\n",
            "285 Video: 012_005_005.mp4 Expected: Man Predicted: Man\n",
            "286 Video: 012_006_005.mp4 Expected: Man Predicted: Man\n",
            "287 Video: 012_007_005.mp4 Expected: Man Predicted: Man\n",
            "288 Video: 012_008_005.mp4 Expected: Man Predicted: Man\n",
            "289 Video: 012_009_005.mp4 Expected: Man Predicted: Man\n",
            "290 Video: 012_010_005.mp4 Expected: Man Predicted: Man\n",
            "291 Video: 021_001_005.mp4 Expected: Milk Predicted: Milk\n",
            "292 Video: 021_002_005.mp4 Expected: Milk Predicted: Milk\n",
            "293 Video: 021_003_005.mp4 Expected: Milk Predicted: Milk\n",
            "294 Video: 021_004_005.mp4 Expected: Milk Predicted: Milk\n",
            "295 Video: 021_005_005.mp4 Expected: Milk Predicted: Milk\n",
            "296 \u001b[31mVideo: 021_006_005.mp4 Expected: Milk Predicted: Skimmer\u001b[0m\n",
            "297 Video: 021_007_005.mp4 Expected: Milk Predicted: Milk\n",
            "298 Video: 021_008_005.mp4 Expected: Milk Predicted: Milk\n",
            "299 Video: 021_009_005.mp4 Expected: Milk Predicted: Milk\n",
            "300 Video: 021_010_005.mp4 Expected: Milk Predicted: Milk\n",
            "301 Video: 063_001_005.mp4 Expected: Give Predicted: Give\n",
            "302 Video: 063_002_005.mp4 Expected: Give Predicted: Give\n",
            "303 Video: 063_003_005.mp4 Expected: Give Predicted: Give\n",
            "304 Video: 063_004_005.mp4 Expected: Give Predicted: Give\n",
            "305 Video: 063_005_005.mp4 Expected: Give Predicted: Give\n",
            "306 Video: 063_006_005.mp4 Expected: Give Predicted: Give\n",
            "307 Video: 063_007_005.mp4 Expected: Give Predicted: Give\n",
            "308 Video: 063_008_005.mp4 Expected: Give Predicted: Give\n",
            "309 Video: 063_009_005.mp4 Expected: Give Predicted: Give\n",
            "310 Video: 063_010_005.mp4 Expected: Give Predicted: Give\n",
            "311 Video: 056_001_005.mp4 Expected: Help Predicted: Help\n",
            "312 Video: 056_002_005.mp4 Expected: Help Predicted: Help\n",
            "313 Video: 056_003_005.mp4 Expected: Help Predicted: Help\n",
            "314 Video: 056_004_005.mp4 Expected: Help Predicted: Help\n",
            "315 Video: 056_005_005.mp4 Expected: Help Predicted: Help\n",
            "316 Video: 056_006_005.mp4 Expected: Help Predicted: Help\n",
            "317 Video: 056_007_005.mp4 Expected: Help Predicted: Help\n",
            "318 Video: 056_008_005.mp4 Expected: Help Predicted: Help\n",
            "319 Video: 056_009_005.mp4 Expected: Help Predicted: Help\n",
            "320 Video: 056_010_005.mp4 Expected: Help Predicted: Help\n",
            "321 Video: 006_001_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "322 Video: 006_002_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "323 Video: 006_003_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "324 Video: 006_004_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "325 Video: 006_005_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "326 Video: 006_006_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "327 Video: 006_007_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "328 Video: 006_008_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "329 Video: 006_009_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "330 Video: 006_010_005.mp4 Expected: Light-blue Predicted: Light-blue\n",
            "331 Video: 033_001_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "332 Video: 033_002_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "333 Video: 033_003_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "334 Video: 033_004_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "335 Video: 033_005_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "336 Video: 033_006_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "337 Video: 033_007_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "338 Video: 033_008_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "339 Video: 033_009_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "340 Video: 033_010_005.mp4 Expected: Hungry Predicted: Hungry\n",
            "341 Video: 016_001_005.mp4 Expected: Learn Predicted: Learn\n",
            "342 Video: 016_002_005.mp4 Expected: Learn Predicted: Learn\n",
            "343 Video: 016_003_005.mp4 Expected: Learn Predicted: Learn\n",
            "344 Video: 016_004_005.mp4 Expected: Learn Predicted: Learn\n",
            "345 Video: 016_005_005.mp4 Expected: Learn Predicted: Learn\n",
            "346 Video: 016_006_005.mp4 Expected: Learn Predicted: Learn\n",
            "347 Video: 016_007_005.mp4 Expected: Learn Predicted: Learn\n",
            "348 Video: 016_008_005.mp4 Expected: Learn Predicted: Learn\n",
            "349 Video: 016_009_005.mp4 Expected: Learn Predicted: Learn\n",
            "350 Video: 016_010_005.mp4 Expected: Learn Predicted: Learn\n",
            "351 Video: 034_001_005.mp4 Expected: Map Predicted: Map\n",
            "352 Video: 034_002_005.mp4 Expected: Map Predicted: Map\n",
            "353 Video: 034_003_005.mp4 Expected: Map Predicted: Map\n",
            "354 Video: 034_004_005.mp4 Expected: Map Predicted: Map\n",
            "355 Video: 034_005_005.mp4 Expected: Map Predicted: Map\n",
            "356 Video: 034_006_005.mp4 Expected: Map Predicted: Map\n",
            "357 Video: 034_007_005.mp4 Expected: Map Predicted: Map\n",
            "358 Video: 034_008_005.mp4 Expected: Map Predicted: Map\n",
            "359 Video: 034_009_005.mp4 Expected: Map Predicted: Map\n",
            "360 Video: 034_010_005.mp4 Expected: Map Predicted: Map\n",
            "361 Video: 007_001_005.mp4 Expected: Colors Predicted: Colors\n",
            "362 Video: 007_002_005.mp4 Expected: Colors Predicted: Colors\n",
            "363 Video: 007_003_005.mp4 Expected: Colors Predicted: Colors\n",
            "364 Video: 007_004_005.mp4 Expected: Colors Predicted: Colors\n",
            "365 Video: 007_005_005.mp4 Expected: Colors Predicted: Colors\n",
            "366 Video: 007_006_005.mp4 Expected: Colors Predicted: Colors\n",
            "367 Video: 007_007_005.mp4 Expected: Colors Predicted: Colors\n",
            "368 Video: 007_008_005.mp4 Expected: Colors Predicted: Colors\n",
            "369 Video: 007_009_005.mp4 Expected: Colors Predicted: Colors\n",
            "370 Video: 007_010_005.mp4 Expected: Colors Predicted: Colors\n",
            "371 Video: 023_001_005.mp4 Expected: Food Predicted: Food\n",
            "372 Video: 023_002_005.mp4 Expected: Food Predicted: Food\n",
            "373 Video: 023_003_005.mp4 Expected: Food Predicted: Food\n",
            "374 Video: 023_004_005.mp4 Expected: Food Predicted: Food\n",
            "375 Video: 023_005_005.mp4 Expected: Food Predicted: Food\n",
            "376 Video: 023_006_005.mp4 Expected: Food Predicted: Food\n",
            "377 Video: 023_007_005.mp4 Expected: Food Predicted: Food\n",
            "378 Video: 023_008_005.mp4 Expected: Food Predicted: Food\n",
            "379 Video: 023_009_005.mp4 Expected: Food Predicted: Food\n",
            "380 Video: 023_010_005.mp4 Expected: Food Predicted: Food\n",
            "381 Video: 014_001_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "382 Video: 014_002_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "383 Video: 014_003_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "384 Video: 014_004_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "385 Video: 014_005_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "386 Video: 014_006_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "387 Video: 014_007_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "388 Video: 014_008_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "389 Video: 014_009_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "390 Video: 014_010_005.mp4 Expected: Drawer Predicted: Drawer\n",
            "391 Video: 057_001_005.mp4 Expected: Dance Predicted: Dance\n",
            "392 Video: 057_002_005.mp4 Expected: Dance Predicted: Dance\n",
            "393 Video: 057_003_005.mp4 Expected: Dance Predicted: Dance\n",
            "394 \u001b[31mVideo: 057_004_005.mp4 Expected: Dance Predicted: Appear\u001b[0m\n",
            "395 Video: 057_005_005.mp4 Expected: Dance Predicted: Dance\n",
            "396 Video: 057_006_005.mp4 Expected: Dance Predicted: Dance\n",
            "397 Video: 057_007_005.mp4 Expected: Dance Predicted: Dance\n",
            "398 Video: 057_008_005.mp4 Expected: Dance Predicted: Dance\n",
            "399 Video: 057_009_005.mp4 Expected: Dance Predicted: Dance\n",
            "400 Video: 057_010_005.mp4 Expected: Dance Predicted: Dance\n",
            "401 Video: 010_001_005.mp4 Expected: Enemy Predicted: Enemy\n",
            "402 Video: 010_002_005.mp4 Expected: Enemy Predicted: Enemy\n",
            "403 Video: 010_003_005.mp4 Expected: Enemy Predicted: Enemy\n",
            "404 \u001b[31mVideo: 010_004_005.mp4 Expected: Enemy Predicted: Drawer\u001b[0m\n",
            "405 Video: 010_005_005.mp4 Expected: Enemy Predicted: Enemy\n",
            "406 Video: 010_006_005.mp4 Expected: Enemy Predicted: Enemy\n",
            "407 \u001b[31mVideo: 010_007_005.mp4 Expected: Enemy Predicted: Opaque\u001b[0m\n",
            "408 Video: 010_008_005.mp4 Expected: Enemy Predicted: Enemy\n",
            "409 \u001b[31mVideo: 010_009_005.mp4 Expected: Enemy Predicted: Skimmer\u001b[0m\n",
            "410 \u001b[31mVideo: 010_010_005.mp4 Expected: Enemy Predicted: Red_2\u001b[0m\n",
            "411 Video: 064_001_005.mp4 Expected: Find Predicted: Find\n",
            "412 Video: 064_002_005.mp4 Expected: Find Predicted: Find\n",
            "413 Video: 064_003_005.mp4 Expected: Find Predicted: Find\n",
            "414 Video: 064_004_005.mp4 Expected: Find Predicted: Find\n",
            "415 Video: 064_005_005.mp4 Expected: Find Predicted: Find\n",
            "416 \u001b[31mVideo: 064_006_005.mp4 Expected: Find Predicted: Hungry\u001b[0m\n",
            "417 Video: 064_007_005.mp4 Expected: Find Predicted: Find\n",
            "418 Video: 064_008_005.mp4 Expected: Find Predicted: Find\n",
            "419 Video: 064_009_005.mp4 Expected: Find Predicted: Find\n",
            "420 Video: 064_010_005.mp4 Expected: Find Predicted: Find\n",
            "421 Video: 035_001_005.mp4 Expected: Coin Predicted: Coin\n",
            "422 Video: 035_002_005.mp4 Expected: Coin Predicted: Coin\n",
            "423 Video: 035_003_005.mp4 Expected: Coin Predicted: Coin\n",
            "424 Video: 035_004_005.mp4 Expected: Coin Predicted: Coin\n",
            "425 Video: 035_005_005.mp4 Expected: Coin Predicted: Coin\n",
            "426 Video: 035_006_005.mp4 Expected: Coin Predicted: Coin\n",
            "427 Video: 035_007_005.mp4 Expected: Coin Predicted: Coin\n",
            "428 Video: 035_008_005.mp4 Expected: Coin Predicted: Coin\n",
            "429 Video: 035_009_005.mp4 Expected: Coin Predicted: Coin\n",
            "430 Video: 035_010_005.mp4 Expected: Coin Predicted: Coin\n",
            "431 Video: 042_001_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "432 Video: 042_002_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "433 Video: 042_003_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "434 Video: 042_004_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "435 Video: 042_005_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "436 Video: 042_006_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "437 Video: 042_007_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "438 Video: 042_008_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "439 Video: 042_009_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "440 Video: 042_010_005.mp4 Expected: Deaf Predicted: Deaf\n",
            "441 Video: 026_001_005.mp4 Expected: Country Predicted: Country\n",
            "442 \u001b[31mVideo: 026_002_005.mp4 Expected: Country Predicted: Deaf\u001b[0m\n",
            "443 Video: 026_003_005.mp4 Expected: Country Predicted: Country\n",
            "444 Video: 026_004_005.mp4 Expected: Country Predicted: Country\n",
            "445 Video: 026_005_005.mp4 Expected: Country Predicted: Country\n",
            "446 Video: 026_006_005.mp4 Expected: Country Predicted: Country\n",
            "447 Video: 026_007_005.mp4 Expected: Country Predicted: Country\n",
            "448 Video: 026_008_005.mp4 Expected: Country Predicted: Country\n",
            "449 Video: 026_009_005.mp4 Expected: Country Predicted: Country\n",
            "450 Video: 026_010_005.mp4 Expected: Country Predicted: Country\n",
            "451 Video: 060_001_005.mp4 Expected: Copy Predicted: Copy\n",
            "452 Video: 060_002_005.mp4 Expected: Copy Predicted: Copy\n",
            "453 Video: 060_003_005.mp4 Expected: Copy Predicted: Copy\n",
            "454 Video: 060_004_005.mp4 Expected: Copy Predicted: Copy\n",
            "455 Video: 060_005_005.mp4 Expected: Copy Predicted: Copy\n",
            "456 Video: 060_006_005.mp4 Expected: Copy Predicted: Copy\n",
            "457 Video: 060_007_005.mp4 Expected: Copy Predicted: Copy\n",
            "458 Video: 060_008_005.mp4 Expected: Copy Predicted: Copy\n",
            "459 Video: 060_009_005.mp4 Expected: Copy Predicted: Copy\n",
            "460 Video: 060_010_005.mp4 Expected: Copy Predicted: Copy\n",
            "461 Video: 019_001_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "462 \u001b[31mVideo: 019_002_005.mp4 Expected: Bitter Predicted: Opaque\u001b[0m\n",
            "463 Video: 019_003_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "464 Video: 019_004_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "465 Video: 019_005_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "466 \u001b[31mVideo: 019_006_005.mp4 Expected: Bitter Predicted: Bright\u001b[0m\n",
            "467 Video: 019_007_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "468 Video: 019_008_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "469 Video: 019_009_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "470 Video: 019_010_005.mp4 Expected: Bitter Predicted: Bitter\n",
            "471 Video: 015_001_005.mp4 Expected: Born Predicted: Born\n",
            "472 Video: 015_002_005.mp4 Expected: Born Predicted: Born\n",
            "473 Video: 015_003_005.mp4 Expected: Born Predicted: Born\n",
            "474 Video: 015_004_005.mp4 Expected: Born Predicted: Born\n",
            "475 Video: 015_005_005.mp4 Expected: Born Predicted: Born\n",
            "476 Video: 015_006_005.mp4 Expected: Born Predicted: Born\n",
            "477 Video: 015_007_005.mp4 Expected: Born Predicted: Born\n",
            "478 Video: 015_008_005.mp4 Expected: Born Predicted: Born\n",
            "479 Video: 015_009_005.mp4 Expected: Born Predicted: Born\n",
            "480 Video: 015_010_005.mp4 Expected: Born Predicted: Born\n",
            "481 Video: 005_001_005.mp4 Expected: Bright Predicted: Bright\n",
            "482 \u001b[31mVideo: 005_002_005.mp4 Expected: Bright Predicted: Away\u001b[0m\n",
            "483 Video: 005_003_005.mp4 Expected: Bright Predicted: Bright\n",
            "484 Video: 005_004_005.mp4 Expected: Bright Predicted: Bright\n",
            "485 Video: 005_005_005.mp4 Expected: Bright Predicted: Bright\n",
            "486 Video: 005_006_005.mp4 Expected: Bright Predicted: Bright\n",
            "487 Video: 005_007_005.mp4 Expected: Bright Predicted: Bright\n",
            "488 Video: 005_008_005.mp4 Expected: Bright Predicted: Bright\n",
            "489 \u001b[31mVideo: 005_009_005.mp4 Expected: Bright Predicted: Skimmer\u001b[0m\n",
            "490 Video: 005_010_005.mp4 Expected: Bright Predicted: Bright\n",
            "491 Video: 031_001_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "492 Video: 031_002_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "493 Video: 031_003_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "494 Video: 031_004_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "495 Video: 031_005_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "496 Video: 031_006_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "497 Video: 031_007_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "498 Video: 031_008_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "499 Video: 031_009_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "500 Video: 031_010_005.mp4 Expected: Breakfast Predicted: Breakfast\n",
            "501 Video: 046_001_005.mp4 Expected: Candy Predicted: Candy\n",
            "502 Video: 046_002_005.mp4 Expected: Candy Predicted: Candy\n",
            "503 Video: 046_003_005.mp4 Expected: Candy Predicted: Candy\n",
            "504 Video: 046_004_005.mp4 Expected: Candy Predicted: Candy\n",
            "505 Video: 046_005_005.mp4 Expected: Candy Predicted: Candy\n",
            "506 Video: 046_006_005.mp4 Expected: Candy Predicted: Candy\n",
            "507 \u001b[31mVideo: 046_007_005.mp4 Expected: Candy Predicted: Deaf\u001b[0m\n",
            "508 Video: 046_008_005.mp4 Expected: Candy Predicted: Candy\n",
            "509 Video: 046_009_005.mp4 Expected: Candy Predicted: Candy\n",
            "510 Video: 046_010_005.mp4 Expected: Candy Predicted: Candy\n",
            "511 Video: 030_001_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "512 Video: 030_002_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "513 Video: 030_003_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "514 Video: 030_004_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "515 Video: 030_005_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "516 Video: 030_006_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "517 Video: 030_007_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "518 Video: 030_008_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "519 Video: 030_009_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "520 Video: 030_010_005.mp4 Expected: Birthday Predicted: Birthday\n",
            "521 Video: 055_001_005.mp4 Expected: Catch Predicted: Catch\n",
            "522 Video: 055_002_005.mp4 Expected: Catch Predicted: Catch\n",
            "523 Video: 055_003_005.mp4 Expected: Catch Predicted: Catch\n",
            "524 Video: 055_004_005.mp4 Expected: Catch Predicted: Catch\n",
            "525 Video: 055_005_005.mp4 Expected: Catch Predicted: Catch\n",
            "526 Video: 055_006_005.mp4 Expected: Catch Predicted: Catch\n",
            "527 Video: 055_007_005.mp4 Expected: Catch Predicted: Catch\n",
            "528 Video: 055_008_005.mp4 Expected: Catch Predicted: Catch\n",
            "529 Video: 055_009_005.mp4 Expected: Catch Predicted: Catch\n",
            "530 Video: 055_010_005.mp4 Expected: Catch Predicted: Catch\n",
            "531 Video: 059_001_005.mp4 Expected: Buy Predicted: Buy\n",
            "532 Video: 059_002_005.mp4 Expected: Buy Predicted: Buy\n",
            "533 Video: 059_003_005.mp4 Expected: Buy Predicted: Buy\n",
            "534 Video: 059_004_005.mp4 Expected: Buy Predicted: Buy\n",
            "535 Video: 059_005_005.mp4 Expected: Buy Predicted: Buy\n",
            "536 Video: 059_006_005.mp4 Expected: Buy Predicted: Buy\n",
            "537 Video: 059_007_005.mp4 Expected: Buy Predicted: Buy\n",
            "538 Video: 059_008_005.mp4 Expected: Buy Predicted: Buy\n",
            "539 Video: 059_009_005.mp4 Expected: Buy Predicted: Buy\n",
            "540 Video: 059_010_005.mp4 Expected: Buy Predicted: Buy\n",
            "541 Video: 017_001_005.mp4 Expected: Call Predicted: Call\n",
            "542 Video: 017_002_005.mp4 Expected: Call Predicted: Call\n",
            "543 Video: 017_003_005.mp4 Expected: Call Predicted: Call\n",
            "544 Video: 017_004_005.mp4 Expected: Call Predicted: Call\n",
            "545 Video: 017_005_005.mp4 Expected: Call Predicted: Call\n",
            "546 Video: 017_006_005.mp4 Expected: Call Predicted: Call\n",
            "547 Video: 017_007_005.mp4 Expected: Call Predicted: Call\n",
            "548 Video: 017_008_005.mp4 Expected: Call Predicted: Call\n",
            "549 Video: 017_009_005.mp4 Expected: Call Predicted: Call\n",
            "550 Video: 017_010_005.mp4 Expected: Call Predicted: Call\n",
            "551 Video: 047_001_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "552 Video: 047_002_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "553 Video: 047_003_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "554 Video: 047_004_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "555 Video: 047_005_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "556 Video: 047_006_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "557 Video: 047_007_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "558 Video: 047_008_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "559 Video: 047_009_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "560 Video: 047_010_005.mp4 Expected: Chewing-gum Predicted: Chewing-gum\n",
            "561 Video: 053_001_005.mp4 Expected: Appear Predicted: Appear\n",
            "562 Video: 053_002_005.mp4 Expected: Appear Predicted: Appear\n",
            "563 Video: 053_003_005.mp4 Expected: Appear Predicted: Appear\n",
            "564 Video: 053_004_005.mp4 Expected: Appear Predicted: Appear\n",
            "565 Video: 053_005_005.mp4 Expected: Appear Predicted: Appear\n",
            "566 Video: 053_006_005.mp4 Expected: Appear Predicted: Appear\n",
            "567 Video: 053_007_005.mp4 Expected: Appear Predicted: Appear\n",
            "568 Video: 053_008_005.mp4 Expected: Appear Predicted: Appear\n",
            "569 Video: 053_009_005.mp4 Expected: Appear Predicted: Appear\n",
            "570 Video: 053_010_005.mp4 Expected: Appear Predicted: Appear\n",
            "571 Video: 024_001_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "572 Video: 024_002_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "573 Video: 024_003_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "574 Video: 024_004_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "575 Video: 024_005_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "576 Video: 024_006_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "577 Video: 024_007_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "578 Video: 024_008_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "579 Video: 024_009_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "580 Video: 024_010_005.mp4 Expected: Argentina Predicted: Argentina\n",
            "581 Video: 058_001_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "582 Video: 058_002_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "583 Video: 058_003_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "584 Video: 058_004_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "585 Video: 058_005_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "586 Video: 058_006_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "587 Video: 058_007_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "588 Video: 058_008_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "589 Video: 058_009_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "590 Video: 058_010_005.mp4 Expected: Bathe Predicted: Bathe\n",
            "591 Video: 045_001_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "592 Video: 045_002_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "593 Video: 045_003_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "594 Video: 045_004_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "595 Video: 045_005_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "596 Video: 045_006_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "597 Video: 045_007_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "598 Video: 045_008_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "599 Video: 045_009_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "600 Video: 045_010_005.mp4 Expected: Barbecue Predicted: Barbecue\n",
            "601 Video: 050_001_005.mp4 Expected: Accept Predicted: Accept\n",
            "602 Video: 050_002_005.mp4 Expected: Accept Predicted: Accept\n",
            "603 Video: 050_003_005.mp4 Expected: Accept Predicted: Accept\n",
            "604 Video: 050_004_005.mp4 Expected: Accept Predicted: Accept\n",
            "605 Video: 050_005_005.mp4 Expected: Accept Predicted: Accept\n",
            "606 Video: 050_006_005.mp4 Expected: Accept Predicted: Accept\n",
            "607 Video: 050_007_005.mp4 Expected: Accept Predicted: Accept\n",
            "608 Video: 050_008_005.mp4 Expected: Accept Predicted: Accept\n",
            "609 Video: 050_009_005.mp4 Expected: Accept Predicted: Accept\n",
            "610 Video: 050_010_005.mp4 Expected: Accept Predicted: Accept\n",
            "611 Video: 013_001_005.mp4 Expected: Away Predicted: Away\n",
            "612 Video: 013_002_005.mp4 Expected: Away Predicted: Away\n",
            "613 Video: 013_003_005.mp4 Expected: Away Predicted: Away\n",
            "614 Video: 013_004_005.mp4 Expected: Away Predicted: Away\n",
            "615 Video: 013_005_005.mp4 Expected: Away Predicted: Away\n",
            "616 Video: 013_006_005.mp4 Expected: Away Predicted: Away\n",
            "617 Video: 013_007_005.mp4 Expected: Away Predicted: Away\n",
            "618 Video: 013_008_005.mp4 Expected: Away Predicted: Away\n",
            "619 \u001b[31mVideo: 013_009_005.mp4 Expected: Away Predicted: Opaque\u001b[0m\n",
            "620 Video: 013_010_005.mp4 Expected: Away Predicted: Away\n",
            "621 Video: 008_001_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "622 \u001b[31mVideo: 008_002_005.mp4 Expected: Red_2 Predicted: Women\u001b[0m\n",
            "623 Video: 008_003_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "624 Video: 008_004_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "625 Video: 008_005_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "626 Video: 008_006_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "627 Video: 008_007_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "628 Video: 008_008_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "629 Video: 008_009_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "630 Video: 008_010_005.mp4 Expected: Red_2 Predicted: Red_2\n",
            "631 Video: 027_001_005.mp4 Expected: Last name Predicted: Last name\n",
            "632 Video: 027_002_005.mp4 Expected: Last name Predicted: Last name\n",
            "633 Video: 027_003_005.mp4 Expected: Last name Predicted: Last name\n",
            "634 Video: 027_004_005.mp4 Expected: Last name Predicted: Last name\n",
            "635 Video: 027_005_005.mp4 Expected: Last name Predicted: Last name\n",
            "636 Video: 027_006_005.mp4 Expected: Last name Predicted: Last name\n",
            "637 Video: 027_007_005.mp4 Expected: Last name Predicted: Last name\n",
            "638 Video: 027_008_005.mp4 Expected: Last name Predicted: Last name\n",
            "639 Video: 027_009_005.mp4 Expected: Last name Predicted: Last name\n",
            "640 Video: 027_010_005.mp4 Expected: Last name Predicted: Last name\n",
            "\n",
            "Accuray: 612/640 , 0.95625\n",
            "time: 1634.5940599441528\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IVMwsF3WchUj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 730
        },
        "outputId": "dfc909e6-90d0-4bd5-bc96-15500e2215e8"
      },
      "source": [
        "sorted(res)\n",
        "ticks=range(len(res))\n",
        "# nếu phần code tiếp theo bị lỗi là do đang sử dụng hàm predictOneByOneReturn2dDict nha\n",
        "height=list(res.values())\n",
        "tick_label=list(res.keys())\n",
        "\n",
        "fig = plt.figure(figsize=(12, 12))\n",
        "plt.barh(ticks,height,tick_label=tick_label,color=['b','m'],height=0.5, align='center')\n",
        "plt.title(\"Accuray on Test\")\n",
        "plt.ylabel(\"Word\")\n",
        "plt.xlabel(\"Number right / total\")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pZ0TsPK9LW0M",
        "outputId": "f9f80b94-a99b-4d78-c50d-82a778ddff8d"
      },
      "source": [
        "a = np.array(list(res.values()))\n",
        "print(a)\n",
        "print(list(res.keys()))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[{'Yogurt': 10} {'Women': 9, 'Drawer': 1} {'Where': 10} {'Trap': 10}\n",
            " {'Uruguay': 10} {'Yellow': 10} {'Water': 10} {'Thanks': 10}\n",
            " {'Bright': 2, 'Milk': 1, 'Skimmer': 7} {'Ship': 10} {'To land': 10}\n",
            " {'Spaghetti': 10} {'Son': 9, 'Light-blue': 1} {'Rice': 10} {'Run': 10}\n",
            " {'Sweet milk': 10} {'Shut down': 8, 'Uruguay': 1, 'Deaf': 1}\n",
            " {'Music': 10} {'Perfume': 10} {'Mock': 10} {'Name': 10} {'Photo': 10}\n",
            " {'Patience': 10} {'Opaque': 8, 'Bright': 1, 'Away': 1} {'None': 10}\n",
            " {'Red': 8, 'Green': 2} {'Realize': 10} {'Green': 8, 'Red': 2} {'Man': 10}\n",
            " {'Milk': 9, 'Skimmer': 1} {'Give': 10} {'Help': 10} {'Light-blue': 10}\n",
            " {'Hungry': 10} {'Learn': 10} {'Map': 10} {'Colors': 10} {'Food': 10}\n",
            " {'Drawer': 10} {'Dance': 9, 'Appear': 1}\n",
            " {'Enemy': 6, 'Drawer': 1, 'Opaque': 1, 'Skimmer': 1, 'Red_2': 1}\n",
            " {'Find': 9, 'Hungry': 1} {'Coin': 10} {'Deaf': 10}\n",
            " {'Country': 9, 'Deaf': 1} {'Copy': 10}\n",
            " {'Bitter': 8, 'Opaque': 1, 'Bright': 1} {'Born': 10}\n",
            " {'Bright': 8, 'Away': 1, 'Skimmer': 1} {'Breakfast': 10}\n",
            " {'Candy': 9, 'Deaf': 1} {'Birthday': 10} {'Catch': 10} {'Buy': 10}\n",
            " {'Call': 10} {'Chewing-gum': 10} {'Appear': 10} {'Argentina': 10}\n",
            " {'Bathe': 10} {'Barbecue': 10} {'Accept': 10} {'Away': 9, 'Opaque': 1}\n",
            " {'Red_2': 9, 'Women': 1} {'Last name': 10}]\n",
            "['Yogurt', 'Women', 'Where', 'Trap', 'Uruguay', 'Yellow', 'Water', 'Thanks', 'Skimmer', 'Ship', 'To land', 'Spaghetti', 'Son', 'Rice', 'Run', 'Sweet milk', 'Shut down', 'Music', 'Perfume', 'Mock', 'Name', 'Photo', 'Patience', 'Opaque', 'None', 'Red', 'Realize', 'Green', 'Man', 'Milk', 'Give', 'Help', 'Light-blue', 'Hungry', 'Learn', 'Map', 'Colors', 'Food', 'Drawer', 'Dance', 'Enemy', 'Find', 'Coin', 'Deaf', 'Country', 'Copy', 'Bitter', 'Born', 'Bright', 'Breakfast', 'Candy', 'Birthday', 'Catch', 'Buy', 'Call', 'Chewing-gum', 'Appear', 'Argentina', 'Bathe', 'Barbecue', 'Accept', 'Away', 'Red_2', 'Last name']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CFJDkLMT9IER",
        "outputId": "31cd2478-9f40-4208-c1fe-7e6ad7ca6499"
      },
      "source": [
        "# exchange result dict to confusion array\n",
        "a = np.zeros((64,64), dtype=int)\n",
        "a_true = list(res.keys())\n",
        "a_pred = list(res.keys())\n",
        "print(a_true)\n",
        "\n",
        "matrix = list(res.values())\n",
        "for w_class in a_true:\n",
        "  for word in list(res[w_class]):\n",
        "    a[a_true.index(w_class)][a_true.index(word)] = res[w_class][word]\n",
        "\n",
        "print(a)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['Yogurt', 'Women', 'Where', 'Trap', 'Uruguay', 'Yellow', 'Water', 'Thanks', 'Skimmer', 'Ship', 'To land', 'Spaghetti', 'Son', 'Rice', 'Run', 'Sweet milk', 'Shut down', 'Music', 'Perfume', 'Mock', 'Name', 'Photo', 'Patience', 'Opaque', 'None', 'Red', 'Realize', 'Green', 'Man', 'Milk', 'Give', 'Help', 'Light-blue', 'Hungry', 'Learn', 'Map', 'Colors', 'Food', 'Drawer', 'Dance', 'Enemy', 'Find', 'Coin', 'Deaf', 'Country', 'Copy', 'Bitter', 'Born', 'Bright', 'Breakfast', 'Candy', 'Birthday', 'Catch', 'Buy', 'Call', 'Chewing-gum', 'Appear', 'Argentina', 'Bathe', 'Barbecue', 'Accept', 'Away', 'Red_2', 'Last name']\n",
            "[[10  0  0 ...  0  0  0]\n",
            " [ 0  9  0 ...  0  0  0]\n",
            " [ 0  0 10 ...  0  0  0]\n",
            " ...\n",
            " [ 0  0  0 ...  9  0  0]\n",
            " [ 0  1  0 ...  0  9  0]\n",
            " [ 0  0  0 ...  0  0 10]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "W8L-fwATKUsS",
        "outputId": "92e0bae6-0c03-432a-f4ef-31b912fc1e67"
      },
      "source": [
        "# remove class with 100% predicted from confusion array\n",
        "i=0\n",
        "while i < (np.shape(a)[0]):\n",
        "    c_true = 0\n",
        "    for j in range(np.shape(a)[1]):\n",
        "        if a[i][j] > 0:\n",
        "            c_true += 1\n",
        "    if c_true == 1:\n",
        "        a = np.delete(a,i,0)\n",
        "        a_true = np.delete(a_true,i,0)\n",
        "        i-=1\n",
        "    i+=1\n",
        "\n",
        "i=0\n",
        "while i < (np.shape(a)[1]):\n",
        "    c_pred = 0\n",
        "    for j in range(np.shape(a)[0]):\n",
        "        if a[j][i] > 0:\n",
        "            c_pred += 1\n",
        "    if c_pred == 0:\n",
        "        a = np.delete(a,i,1)\n",
        "        a_pred = np.delete(a_pred,i,0)\n",
        "        i-=1\n",
        "    i+=1\n",
        "\n",
        "print('a_true: ',a_true)\n",
        "print('a_pred: ',a_pred)\n",
        "print(a)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "a_true:  ['Women' 'Skimmer' 'Son' 'Shut down' 'Opaque' 'Red' 'Green' 'Milk' 'Dance'\n",
            " 'Enemy' 'Find' 'Country' 'Bitter' 'Bright' 'Candy' 'Away' 'Red_2']\n",
            "a_pred:  ['Women' 'Uruguay' 'Skimmer' 'Son' 'Shut down' 'Opaque' 'Red' 'Green'\n",
            " 'Milk' 'Light-blue' 'Hungry' 'Drawer' 'Dance' 'Enemy' 'Find' 'Deaf'\n",
            " 'Country' 'Bitter' 'Bright' 'Candy' 'Appear' 'Away' 'Red_2']\n",
            "[[9 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0]\n",
            " [0 0 7 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 2 0 0 0 0]\n",
            " [0 0 0 9 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0]\n",
            " [0 1 0 0 8 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0]\n",
            " [0 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0]\n",
            " [0 0 0 0 0 0 8 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]\n",
            " [0 0 0 0 0 0 2 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]\n",
            " [0 0 1 0 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0]\n",
            " [0 0 0 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 0 0 1 0 0]\n",
            " [0 0 1 0 0 1 0 0 0 0 0 1 0 6 0 0 0 0 0 0 0 0 1]\n",
            " [0 0 0 0 0 0 0 0 0 0 1 0 0 0 9 0 0 0 0 0 0 0 0]\n",
            " [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 9 0 0 0 0 0 0]\n",
            " [0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 8 1 0 0 0 0]\n",
            " [0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0 1 0]\n",
            " [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 9 0 0 0]\n",
            " [0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 0]\n",
            " [1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TIP-1IUQsKdt",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 672
        },
        "outputId": "263ba03c-657e-4ab5-ce1c-ce2eac90980d"
      },
      "source": [
        "# draw the confusion matrix\n",
        "import seaborn as sn\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "df_cm = pd.DataFrame(a, index = [i for i in a_true], columns = [i for i in a_pred])\n",
        "plt.figure(figsize=(15,10))\n",
        "sn.set(font_scale=1.4) # for label size\n",
        "sn.heatmap(df_cm, annot=True, cmap = 'Greens') # font size\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p6wBZusEM7wR",
        "outputId": "3ccb7d51-aedc-4f66-b05f-0bd8f0788644"
      },
      "source": [
        "# save the predict result dictionary to csv\n",
        "import csv\n",
        "with open('predicted_directory.csv', 'w') as f:\n",
        "    for w_class in res.keys():\n",
        "      for word in res[w_class]:\n",
        "        f.write(\"%s,%s,%s\\n\"%(w_class,word,res[w_class][word]))\n",
        "        print(w_class, \" \", word, \" \", res[w_class][word])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Yogurt   Yogurt   10\n",
            "Women   Women   9\n",
            "Women   Drawer   1\n",
            "Where   Where   10\n",
            "Trap   Trap   10\n",
            "Uruguay   Uruguay   10\n",
            "Yellow   Yellow   10\n",
            "Water   Water   10\n",
            "Thanks   Thanks   10\n",
            "Skimmer   Bright   2\n",
            "Skimmer   Milk   1\n",
            "Skimmer   Skimmer   7\n",
            "Ship   Ship   10\n",
            "To land   To land   10\n",
            "Spaghetti   Spaghetti   10\n",
            "Son   Son   9\n",
            "Son   Light-blue   1\n",
            "Rice   Rice   10\n",
            "Run   Run   10\n",
            "Sweet milk   Sweet milk   10\n",
            "Shut down   Shut down   8\n",
            "Shut down   Uruguay   1\n",
            "Shut down   Deaf   1\n",
            "Music   Music   10\n",
            "Perfume   Perfume   10\n",
            "Mock   Mock   10\n",
            "Name   Name   10\n",
            "Photo   Photo   10\n",
            "Patience   Patience   10\n",
            "Opaque   Opaque   8\n",
            "Opaque   Bright   1\n",
            "Opaque   Away   1\n",
            "None   None   10\n",
            "Red   Red   8\n",
            "Red   Green   2\n",
            "Realize   Realize   10\n",
            "Green   Green   8\n",
            "Green   Red   2\n",
            "Man   Man   10\n",
            "Milk   Milk   9\n",
            "Milk   Skimmer   1\n",
            "Give   Give   10\n",
            "Help   Help   10\n",
            "Light-blue   Light-blue   10\n",
            "Hungry   Hungry   10\n",
            "Learn   Learn   10\n",
            "Map   Map   10\n",
            "Colors   Colors   10\n",
            "Food   Food   10\n",
            "Drawer   Drawer   10\n",
            "Dance   Dance   9\n",
            "Dance   Appear   1\n",
            "Enemy   Enemy   6\n",
            "Enemy   Drawer   1\n",
            "Enemy   Opaque   1\n",
            "Enemy   Skimmer   1\n",
            "Enemy   Red_2   1\n",
            "Find   Find   9\n",
            "Find   Hungry   1\n",
            "Coin   Coin   10\n",
            "Deaf   Deaf   10\n",
            "Country   Country   9\n",
            "Country   Deaf   1\n",
            "Copy   Copy   10\n",
            "Bitter   Bitter   8\n",
            "Bitter   Opaque   1\n",
            "Bitter   Bright   1\n",
            "Born   Born   10\n",
            "Bright   Bright   8\n",
            "Bright   Away   1\n",
            "Bright   Skimmer   1\n",
            "Breakfast   Breakfast   10\n",
            "Candy   Candy   9\n",
            "Candy   Deaf   1\n",
            "Birthday   Birthday   10\n",
            "Catch   Catch   10\n",
            "Buy   Buy   10\n",
            "Call   Call   10\n",
            "Chewing-gum   Chewing-gum   10\n",
            "Appear   Appear   10\n",
            "Argentina   Argentina   10\n",
            "Bathe   Bathe   10\n",
            "Barbecue   Barbecue   10\n",
            "Accept   Accept   10\n",
            "Away   Away   9\n",
            "Away   Opaque   1\n",
            "Red_2   Red_2   9\n",
            "Red_2   Women   1\n",
            "Last name   Last name   10\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}