fer2013_ensemble_mlcnns_testing_icprai2018.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# This source code belongs to paper:\n",
    "# Facial Expression Recognition Using a Multi-level Convolutional Neural Network\n",
    "\n",
    "# Hai-Duong Nguyen*, Soonja Yeom**, Il-Seok Oh***, Kyoung-Min Kim****, Soo-Hyung Kim*\n",
    "# (*) School of Electronics and Computer Engineering, Chonnam National University, South Korea\n",
    "# (**) School of Engineering and ICT, University of Tasmania, Australia\n",
    "# (***) Department of Computer Science and Engineering, Chonbuk National University, South Korea\n",
    "# (****) Department of Electrical and Semiconductor Engineering, Chonnam National University, South Korea"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import load_model\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import itertools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_classes = 7\n",
    "classes=np.array(('angry', 'disgust', 'fear', 'happy', 'neutral', 'sad', 'surprise'))\n",
    "\n",
    "img_size = 48\n",
    "# batch_size = 1024\n",
    "batch_size = 64\n",
    "\n",
    "current_model = 'fer2013'\n",
    "\n",
    "final_model_path = current_model + '.h5'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Other functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(y_test, y_pred, classes,\n",
    "                          normalize=False,\n",
    "                          title='Unnormalized confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    cm = confusion_matrix(y_test, y_pred)\n",
    "    \n",
    "    if normalize:\n",
    "        cm = np.round(cm.astype('float') / cm.sum(axis=1)[:, np.newaxis], 2)\n",
    "        \n",
    "    np.set_printoptions(precision=2)\n",
    "        \n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    thresh = cm.min() + (cm.max() - cm.min()) / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, cm[i, j],\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True expression')\n",
    "    plt.xlabel('Predicted expression')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def int2emo(num):\n",
    "    if num == 0:\n",
    "        return 'Angry'\n",
    "    elif num == 1:\n",
    "        return 'Disgust'\n",
    "    elif num == 2:\n",
    "        return 'Fear'\n",
    "    elif num == 3:\n",
    "        return 'Happy'\n",
    "    elif num == 4:\n",
    "        return 'Neutral'\n",
    "    elif num == 5:\n",
    "        return 'Sad'\n",
    "    else:\n",
    "        return 'Surprise'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test_data_x = np.load(current_model + '_test_data_x.npy')\n",
    "test_data_y_one_hot = np.load(current_model + '_test_data_y_one_hot.npy')\n",
    "n_test_ins = test_data_x.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MLCNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> loading trained model from ./fer2013.hdf5...done.\n"
     ]
    }
   ],
   "source": [
    "BEST_WEIGHT_FILE = './' + current_model + '.hdf5'\n",
    "\n",
    "print('> loading trained model from ' + BEST_WEIGHT_FILE + '...', end='')\n",
    "model = load_model(BEST_WEIGHT_FILE)\n",
    "print('done.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                     Output Shape          Param #     Connected to                     \n",
      "====================================================================================================\n",
      "input (InputLayer)               (None, 48, 48, 1)     0                                            \n",
      "____________________________________________________________________________________________________\n",
      "conv_1_1_net_1 (Conv2D)          (None, 48, 48, 64)    640         input[0][0]                      \n",
      "____________________________________________________________________________________________________\n",
      "conv_1_1_net_2 (Conv2D)          (None, 48, 48, 64)    640         input[0][0]                      \n",
      "____________________________________________________________________________________________________\n",
      "conv_1_1_net_3 (Conv2D)          (None, 48, 48, 64)    640         input[0][0]                      \n",
      "____________________________________________________________________________________________________\n",
      "conv_1_2_net_1 (Conv2D)          (None, 48, 48, 64)    36928       conv_1_1_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_1_2_net_2 (Conv2D)          (None, 48, 48, 64)    36928       conv_1_1_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_1_2_net_3 (Conv2D)          (None, 48, 48, 64)    36928       conv_1_1_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_1_net_1 (MaxPooling2D)      (None, 24, 24, 64)    0           conv_1_2_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_1_net_2 (MaxPooling2D)      (None, 24, 24, 64)    0           conv_1_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_1_net_3 (MaxPooling2D)      (None, 24, 24, 64)    0           conv_1_2_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "drop_1_net_1 (Dropout)           (None, 24, 24, 64)    0           pool_1_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_1_net_2 (Dropout)           (None, 24, 24, 64)    0           pool_1_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_1_net_3 (Dropout)           (None, 24, 24, 64)    0           pool_1_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_1_net_1 (Conv2D)          (None, 24, 24, 128)   73856       drop_1_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_1_net_2 (Conv2D)          (None, 24, 24, 128)   73856       drop_1_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_1_net_3 (Conv2D)          (None, 24, 24, 128)   73856       drop_1_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_2_net_1 (Conv2D)          (None, 24, 24, 128)   147584      conv_2_1_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_2_net_2 (Conv2D)          (None, 24, 24, 128)   147584      conv_2_1_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_2_net_3 (Conv2D)          (None, 24, 24, 128)   147584      conv_2_1_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_3_net_1 (Conv2D)          (None, 24, 24, 128)   147584      conv_2_2_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_3_net_2 (Conv2D)          (None, 24, 24, 128)   147584      conv_2_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_2_3_net_3 (Conv2D)          (None, 24, 24, 128)   147584      conv_2_2_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_2_net_1 (MaxPooling2D)      (None, 12, 12, 128)   0           conv_2_3_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_2_net_2 (MaxPooling2D)      (None, 12, 12, 128)   0           conv_2_3_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_2_net_3 (MaxPooling2D)      (None, 12, 12, 128)   0           conv_2_3_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "drop_2_net_1 (Dropout)           (None, 12, 12, 128)   0           pool_2_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_2_net_2 (Dropout)           (None, 12, 12, 128)   0           pool_2_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_2_net_3 (Dropout)           (None, 12, 12, 128)   0           pool_2_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_1_net_1 (Conv2D)          (None, 12, 12, 256)   295168      drop_2_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_1_net_2 (Conv2D)          (None, 12, 12, 256)   295168      drop_2_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_1_net_3 (Conv2D)          (None, 12, 12, 256)   295168      drop_2_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_2_net_1 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_1_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_2_net_2 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_1_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_2_net_3 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_1_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_3_net_1 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_2_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_3_net_2 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_3_net_3 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_2_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_4_net_1 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_3_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_4_net_2 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_3_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_3_4_net_3 (Conv2D)          (None, 12, 12, 256)   590080      conv_3_3_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_3_net_1 (MaxPooling2D)      (None, 6, 6, 256)     0           conv_3_4_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_3_net_2 (MaxPooling2D)      (None, 6, 6, 256)     0           conv_3_4_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_3_net_3 (MaxPooling2D)      (None, 6, 6, 256)     0           conv_3_4_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "drop_3_net_1 (Dropout)           (None, 6, 6, 256)     0           pool_3_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_3_net_2 (Dropout)           (None, 6, 6, 256)     0           pool_3_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_3_net_3 (Dropout)           (None, 6, 6, 256)     0           pool_3_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_1_net_1 (Conv2D)          (None, 6, 6, 256)     590080      drop_3_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_1_net_2 (Conv2D)          (None, 6, 6, 256)     590080      drop_3_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_1_net_3 (Conv2D)          (None, 6, 6, 256)     590080      drop_3_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_2_net_1 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_1_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_2_net_2 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_1_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_2_net_3 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_1_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_3_net_1 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_2_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_3_net_2 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_3_net_3 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_2_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_4_net_1 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_3_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_4_net_2 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_3_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_4_4_net_3 (Conv2D)          (None, 6, 6, 256)     590080      conv_4_3_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_4_net_1 (MaxPooling2D)      (None, 3, 3, 256)     0           conv_4_4_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_4_net_2 (MaxPooling2D)      (None, 3, 3, 256)     0           conv_4_4_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_4_net_3 (MaxPooling2D)      (None, 3, 3, 256)     0           conv_4_4_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "drop_4_net_1 (Dropout)           (None, 3, 3, 256)     0           pool_4_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_4_net_2 (Dropout)           (None, 3, 3, 256)     0           pool_4_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_4_net_3 (Dropout)           (None, 3, 3, 256)     0           pool_4_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_1_net_1 (Conv2D)          (None, 3, 3, 512)     1180160     drop_4_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_1_net_2 (Conv2D)          (None, 3, 3, 512)     1180160     drop_4_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_1_net_3 (Conv2D)          (None, 3, 3, 512)     1180160     drop_4_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_2_net_1 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_1_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_2_net_2 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_1_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_2_net_3 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_1_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_3_net_1 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_2_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_3_net_2 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_3_net_3 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_2_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_4_net_1 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_3_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_4_net_2 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_3_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "conv_5_4_net_3 (Conv2D)          (None, 3, 3, 512)     2359808     conv_5_3_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_5_net_1 (MaxPooling2D)      (None, 1, 1, 512)     0           conv_5_4_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_5_net_2 (MaxPooling2D)      (None, 1, 1, 512)     0           conv_5_4_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "pool_5_net_3 (MaxPooling2D)      (None, 1, 1, 512)     0           conv_5_4_net_3[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "flatten_2_net_1 (Flatten)        (None, 73728)         0           conv_2_3_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "flatten_3_net_1 (Flatten)        (None, 36864)         0           conv_3_4_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "flatten_4_net_1 (Flatten)        (None, 9216)          0           conv_4_4_net_1[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "drop_5_net_1 (Dropout)           (None, 1, 1, 512)     0           pool_5_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "flatten_2_net_2 (Flatten)        (None, 36864)         0           conv_3_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "flatten_3_net_2 (Flatten)        (None, 9216)          0           conv_4_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "flatten_4_net_2 (Flatten)        (None, 4608)          0           conv_5_2_net_2[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "drop_5_net_2 (Dropout)           (None, 1, 1, 512)     0           pool_5_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "flatten_2_net_3 (Flatten)        (None, 18432)         0           drop_2_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "flatten_3_net_3 (Flatten)        (None, 9216)          0           drop_3_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "flatten_4_net_3 (Flatten)        (None, 2304)          0           drop_4_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "drop_5_net_3 (Dropout)           (None, 1, 1, 512)     0           pool_5_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "dense_2_net_1 (Dense)            (None, 256)           18874624    flatten_2_net_1[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "dense_3_net_1 (Dense)            (None, 256)           9437440     flatten_3_net_1[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "dense_4_net_1 (Dense)            (None, 256)           2359552     flatten_4_net_1[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "flatten_5_net_1 (Flatten)        (None, 512)           0           drop_5_net_1[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "dense_2_net_2 (Dense)            (None, 256)           9437440     flatten_2_net_2[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "dense_3_net_2 (Dense)            (None, 256)           2359552     flatten_3_net_2[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "dense_4_net_2 (Dense)            (None, 256)           1179904     flatten_4_net_2[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "flatten_5_net_2 (Flatten)        (None, 512)           0           drop_5_net_2[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "dense_2_net_3 (Dense)            (None, 256)           4718848     flatten_2_net_3[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "dense_3_net_3 (Dense)            (None, 256)           2359552     flatten_3_net_3[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "dense_4_net_3 (Dense)            (None, 256)           590080      flatten_4_net_3[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "flatten_5_net_3 (Flatten)        (None, 512)           0           drop_5_net_3[0][0]               \n",
      "____________________________________________________________________________________________________\n",
      "early_confusion_net_1 (Concatena (None, 1280)          0           dense_2_net_1[0][0]              \n",
      "                                                                   dense_3_net_1[0][0]              \n",
      "                                                                   dense_4_net_1[0][0]              \n",
      "                                                                   flatten_5_net_1[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "early_confusion_net_2 (Concatena (None, 1280)          0           dense_2_net_2[0][0]              \n",
      "                                                                   dense_3_net_2[0][0]              \n",
      "                                                                   dense_4_net_2[0][0]              \n",
      "                                                                   flatten_5_net_2[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "early_confusion_net_3 (Concatena (None, 1280)          0           dense_2_net_3[0][0]              \n",
      "                                                                   dense_3_net_3[0][0]              \n",
      "                                                                   dense_4_net_3[0][0]              \n",
      "                                                                   flatten_5_net_3[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "later_confusion (Concatenate)    (None, 3840)          0           early_confusion_net_1[0][0]      \n",
      "                                                                   early_confusion_net_2[0][0]      \n",
      "                                                                   early_confusion_net_3[0][0]      \n",
      "____________________________________________________________________________________________________\n",
      "ensemble_dense_1 (Dense)         (None, 512)           1966592     later_confusion[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "ensemble_drop_1 (Dropout)        (None, 512)           0           ensemble_dense_1[0][0]           \n",
      "____________________________________________________________________________________________________\n",
      "ensemble_dense_2 (Dense)         (None, 512)           262656      ensemble_drop_1[0][0]            \n",
      "____________________________________________________________________________________________________\n",
      "ensemble_drop_2 (Dropout)        (None, 512)           0           ensemble_dense_2[0][0]           \n",
      "____________________________________________________________________________________________________\n",
      "ensemble_7_way_softmax (Dense)   (None, 7)             3591        ensemble_drop_2[0][0]            \n",
      "====================================================================================================\n",
      "Total params: 92,825,543\n",
      "Trainable params: 2,232,839\n",
      "Non-trainable params: 90,592,704\n",
      "____________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files\\Anaconda3\\lib\\site-packages\\keras\\preprocessing\\image.py:653: UserWarning: Expected input to be images (as Numpy array) following the data format convention \"channels_last\" (channels on axis 3), i.e. expected either 1, 3 or 4 channels on axis 3. However, it was passed an array with shape (3589, 48, 48, 1) (1 channels).\n",
      "  ' (' + str(x.shape[self.channel_axis]) + ' channels).')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "57/56 [==============================] - 4s     \n"
     ]
    }
   ],
   "source": [
    "test_datagen = ImageDataGenerator(featurewise_center=True, featurewise_std_normalization=True)\n",
    "test_datagen.fit(test_data_x)\n",
    "test_generator = test_datagen.flow(test_data_x, test_data_y_one_hot, batch_size=batch_size, shuffle=False)\n",
    "\n",
    "y_pred_ = model.predict_generator(generator=test_generator, steps=n_test_ins/batch_size, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAb0AAAGaCAYAAACbh53KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a455d5d748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Predictions\n",
    "y_pred = np.argmax(y_pred_, axis=1)\n",
    "# Ground truth\n",
    "t_te = np.argmax(test_generator.y, axis=1)\n",
    "\n",
    "fig = plot_confusion_matrix(y_test=t_te, y_pred=y_pred,\n",
    "                      classes=classes,\n",
    "                      normalize=True,\n",
    "                      cmap=plt.cm.Greys,\n",
    "                      title='Average accuracy: ' + str(np.sum(y_pred == t_te)/len(t_te)) + '\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Testing with random image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "rnd_indx = np.random.randint(0, test_data_x.shape[0])\n",
    "test_img = test_data_x[rnd_indx, :, :]\n",
    "test_img = test_img.reshape((1, test_img.shape[0], test_img.shape[1], test_img.shape[2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pred_emo = int2emo(y_pred[rnd_indx])\n",
    "true_emo = int2emo(t_te[rnd_indx])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAGcCAYAAAAI8xEQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZQXhVVZZZZU3lP8HtWlWDfgkabAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a455d50e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(test_img[0, :, :, 0], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.title('True emotion: ' + true_emo + '\\nPredicted emotion: ' + pred_emo + '\\n')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}