Emotion_Detector.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "3DR-eO17geWu"
   },
   "source": [
    "# Convolutional Neural Network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "EMefrVPCg-60"
   },
   "source": [
    "### Importing the libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "sCV30xyVhFbE"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "oxQxCBWyoGPE"
   },
   "source": [
    "## Part 1 - Data Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "MvE-heJNo3GG"
   },
   "source": [
    "### Preprocessing the Training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0koUcJMJpEBD"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 7442 images belonging to 6 classes.\n"
     ]
    }
   ],
   "source": [
    "dataset_train = ImageDataGenerator(rescale = 1./255,\n",
    "                                   shear_range = 0.2,\n",
    "                                   zoom_range = 0.2,\n",
    "                                   horizontal_flip = True)\n",
    "training_set = dataset_train.flow_from_directory('D:/Emotion Detector/data/Train',\n",
    "                                                 color_mode=\"grayscale\",\n",
    "                                                 target_size = (50,50),\n",
    "                                                 batch_size = 32 ,\n",
    "                                                 class_mode = 'categorical')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "mrCMmGw9pHys"
   },
   "source": [
    "### Preprocessing the Test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "SH4WzfOhpKc3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 4529 images belonging to 6 classes.\n"
     ]
    }
   ],
   "source": [
    "dataset_test = ImageDataGenerator(rescale = 1./255)\n",
    "test_set = dataset_test.flow_from_directory('D:/Emotion Detector/data/Test',\n",
    "                                                color_mode=\"grayscale\",\n",
    "                                                 target_size = (50,50),\n",
    "                                                 batch_size = 32,\n",
    "                                                 class_mode = 'categorical')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "af8O4l90gk7B"
   },
   "source": [
    "## Part 2 - Building the CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  ...\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.00162191]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  ...\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.00162191]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  ...\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.00162191]]\n",
      "\n",
      " ...\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  ...\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  ...\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  ...\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]]\n"
     ]
    }
   ],
   "source": [
    "print(training_set[0][0][0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ces1gXY2lmoX"
   },
   "source": [
    "### Initialising the CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "SAUt4UMPlhLS"
   },
   "outputs": [],
   "source": [
    "cnn = tf.keras.models.Sequential()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "u5YJj_XMl5LF"
   },
   "source": [
    "### Step 1 - Convolution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XPzPrMckl-hV"
   },
   "outputs": [],
   "source": [
    "cnn.add(tf.keras.layers.Conv2D(filters=32,kernel_size=3,activation='relu',input_shape=[50,50,1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "tf87FpvxmNOJ"
   },
   "source": [
    "### Step 2 - Pooling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ncpqPl69mOac"
   },
   "outputs": [],
   "source": [
    "cnn.add(tf.keras.layers.MaxPool2D(pool_size=2,strides=2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xaTOgD8rm4mU"
   },
   "source": [
    "### Adding a second convolutional layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "i_-FZjn_m8gk"
   },
   "outputs": [],
   "source": [
    "cnn.add(tf.keras.layers.Conv2D(filters=32,kernel_size=3,activation='relu'))\n",
    "cnn.add(tf.keras.layers.MaxPool2D(pool_size=2,strides=2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "tmiEuvTunKfk"
   },
   "source": [
    "### Step 3 - Flattening"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "6AZeOGCvnNZn"
   },
   "outputs": [],
   "source": [
    "cnn.add(tf.keras.layers.Flatten())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "dAoSECOm203v"
   },
   "source": [
    "### Step 4 - Full Connection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "8GtmUlLd26Nq"
   },
   "outputs": [],
   "source": [
    "cnn.add(tf.keras.layers.Dense(units=512,activation='relu'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "yTldFvbX28Na"
   },
   "source": [
    "### Step 5 - Output Layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1p_Zj1Mc3Ko_"
   },
   "outputs": [],
   "source": [
    "cnn.add(tf.keras.layers.Dense(units=6,activation='softmax'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "D6XkI90snSDl"
   },
   "source": [
    "## Part 3 - Training the CNN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "vfrFQACEnc6i"
   },
   "source": [
    "### Compiling the CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "NALksrNQpUlJ"
   },
   "outputs": [],
   "source": [
    "cnn.compile(optimizer = 'adam',loss = 'categorical_crossentropy', metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_5\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_10 (Conv2D)           (None, 48, 48, 32)        320       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_10 (MaxPooling (None, 24, 24, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_11 (Conv2D)           (None, 22, 22, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_11 (MaxPooling (None, 11, 11, 32)        0         \n",
      "_________________________________________________________________\n",
      "flatten_5 (Flatten)          (None, 3872)              0         \n",
      "_________________________________________________________________\n",
      "dense_10 (Dense)             (None, 512)               1982976   \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 6)                 3078      \n",
      "=================================================================\n",
      "Total params: 1,995,622\n",
      "Trainable params: 1,995,622\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "cnn.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ehS-v3MIpX2h"
   },
   "source": [
    "### Training the CNN on the Training set and evaluating it on the Test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XUj1W4PJptta",
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/70\n",
      "233/233 [==============================] - 14s 61ms/step - loss: 1.6768 - accuracy: 0.2648 - val_loss: 1.5557 - val_accuracy: 0.3595\n",
      "Epoch 2/70\n",
      "233/233 [==============================] - 14s 61ms/step - loss: 1.5588 - accuracy: 0.3452 - val_loss: 1.5162 - val_accuracy: 0.3935\n",
      "Epoch 3/70\n",
      "233/233 [==============================] - 14s 61ms/step - loss: 1.5072 - accuracy: 0.3733 - val_loss: 1.5464 - val_accuracy: 0.3937\n",
      "Epoch 4/70\n",
      "233/233 [==============================] - 14s 61ms/step - loss: 1.4487 - accuracy: 0.4028 - val_loss: 1.4234 - val_accuracy: 0.4052\n",
      "Epoch 5/70\n",
      "233/233 [==============================] - 14s 61ms/step - loss: 1.4256 - accuracy: 0.4211 - val_loss: 1.4209 - val_accuracy: 0.4200\n",
      "Epoch 6/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 1.3893 - accuracy: 0.4351 - val_loss: 1.4065 - val_accuracy: 0.4509\n",
      "Epoch 7/70\n",
      "233/233 [==============================] - 14s 61ms/step - loss: 1.3690 - accuracy: 0.4429 - val_loss: 1.3446 - val_accuracy: 0.4747\n",
      "Epoch 8/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 1.3512 - accuracy: 0.4592 - val_loss: 1.4113 - val_accuracy: 0.4348\n",
      "Epoch 9/70\n",
      "233/233 [==============================] - 15s 65ms/step - loss: 1.3453 - accuracy: 0.4558 - val_loss: 1.4476 - val_accuracy: 0.4001\n",
      "Epoch 10/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 1.3087 - accuracy: 0.4794 - val_loss: 1.2109 - val_accuracy: 0.5184\n",
      "Epoch 11/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 1.2838 - accuracy: 0.4862 - val_loss: 1.3357 - val_accuracy: 0.4696\n",
      "Epoch 12/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 1.2903 - accuracy: 0.4897 - val_loss: 1.2740 - val_accuracy: 0.4911\n",
      "Epoch 13/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 1.2644 - accuracy: 0.4970 - val_loss: 1.1945 - val_accuracy: 0.5286\n",
      "Epoch 14/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 1.2428 - accuracy: 0.5069 - val_loss: 1.1583 - val_accuracy: 0.5440\n",
      "Epoch 15/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 1.2289 - accuracy: 0.5095 - val_loss: 1.2201 - val_accuracy: 0.5083\n",
      "Epoch 16/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.2194 - accuracy: 0.5120 - val_loss: 1.0873 - val_accuracy: 0.5783\n",
      "Epoch 17/70\n",
      "233/233 [==============================] - 16s 67ms/step - loss: 1.2049 - accuracy: 0.5228 - val_loss: 1.0920 - val_accuracy: 0.5690\n",
      "Epoch 18/70\n",
      "233/233 [==============================] - 15s 66ms/step - loss: 1.1809 - accuracy: 0.5308 - val_loss: 1.1048 - val_accuracy: 0.5648\n",
      "Epoch 19/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 1.1576 - accuracy: 0.5419 - val_loss: 1.0846 - val_accuracy: 0.5697\n",
      "Epoch 20/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 1.1462 - accuracy: 0.5447 - val_loss: 1.0285 - val_accuracy: 0.5997\n",
      "Epoch 21/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.1425 - accuracy: 0.5466 - val_loss: 1.0557 - val_accuracy: 0.5789\n",
      "Epoch 22/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.1178 - accuracy: 0.5636 - val_loss: 1.0672 - val_accuracy: 0.5741\n",
      "Epoch 23/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 1.1013 - accuracy: 0.5688 - val_loss: 0.9579 - val_accuracy: 0.6330\n",
      "Epoch 24/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 1.0809 - accuracy: 0.5822 - val_loss: 0.9693 - val_accuracy: 0.6191\n",
      "Epoch 25/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.0609 - accuracy: 0.5920 - val_loss: 0.9403 - val_accuracy: 0.6441\n",
      "Epoch 26/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.0510 - accuracy: 0.5950 - val_loss: 0.9080 - val_accuracy: 0.6533\n",
      "Epoch 27/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.0331 - accuracy: 0.5985 - val_loss: 0.9445 - val_accuracy: 0.6357\n",
      "Epoch 28/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 1.0278 - accuracy: 0.5992 - val_loss: 0.8945 - val_accuracy: 0.6578\n",
      "Epoch 29/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.9996 - accuracy: 0.6148 - val_loss: 0.8779 - val_accuracy: 0.6666\n",
      "Epoch 30/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 0.9742 - accuracy: 0.6275 - val_loss: 0.8762 - val_accuracy: 0.6681\n",
      "Epoch 31/70\n",
      "233/233 [==============================] - 15s 64ms/step - loss: 0.9671 - accuracy: 0.6344 - val_loss: 0.9044 - val_accuracy: 0.6516\n",
      "Epoch 32/70\n",
      "233/233 [==============================] - 15s 64ms/step - loss: 0.9552 - accuracy: 0.6303 - val_loss: 0.8399 - val_accuracy: 0.6814\n",
      "Epoch 33/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.9429 - accuracy: 0.6422 - val_loss: 0.8213 - val_accuracy: 0.6911\n",
      "Epoch 34/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.9130 - accuracy: 0.6549 - val_loss: 0.8125 - val_accuracy: 0.6865\n",
      "Epoch 35/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.9069 - accuracy: 0.6518 - val_loss: 0.7255 - val_accuracy: 0.7311\n",
      "Epoch 36/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.8761 - accuracy: 0.6666 - val_loss: 0.7100 - val_accuracy: 0.7353\n",
      "Epoch 37/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.8703 - accuracy: 0.6720 - val_loss: 0.7612 - val_accuracy: 0.7094\n",
      "Epoch 38/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.8422 - accuracy: 0.6766 - val_loss: 0.7954 - val_accuracy: 0.6957\n",
      "Epoch 39/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.8286 - accuracy: 0.6872 - val_loss: 0.7094 - val_accuracy: 0.7322\n",
      "Epoch 40/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7923 - accuracy: 0.7006 - val_loss: 0.6166 - val_accuracy: 0.7706\n",
      "Epoch 41/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7957 - accuracy: 0.6962 - val_loss: 0.6009 - val_accuracy: 0.7788\n",
      "Epoch 42/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7730 - accuracy: 0.7076 - val_loss: 0.6088 - val_accuracy: 0.7726\n",
      "Epoch 43/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7529 - accuracy: 0.7184 - val_loss: 0.6053 - val_accuracy: 0.7730\n",
      "Epoch 44/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7369 - accuracy: 0.7221 - val_loss: 0.6030 - val_accuracy: 0.7735\n",
      "Epoch 45/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7367 - accuracy: 0.7210 - val_loss: 0.6327 - val_accuracy: 0.7609\n",
      "Epoch 46/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.7096 - accuracy: 0.7361 - val_loss: 0.5393 - val_accuracy: 0.7951\n",
      "Epoch 47/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.6991 - accuracy: 0.7385 - val_loss: 0.5583 - val_accuracy: 0.7898\n",
      "Epoch 48/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.6827 - accuracy: 0.7487 - val_loss: 0.6212 - val_accuracy: 0.7582\n",
      "Epoch 49/70\n",
      "233/233 [==============================] - 15s 65ms/step - loss: 0.6822 - accuracy: 0.7466 - val_loss: 0.6024 - val_accuracy: 0.7735\n",
      "Epoch 50/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.6567 - accuracy: 0.7517 - val_loss: 0.4664 - val_accuracy: 0.8351\n",
      "Epoch 51/70\n",
      "233/233 [==============================] - 15s 65ms/step - loss: 0.6527 - accuracy: 0.7580 - val_loss: 0.5376 - val_accuracy: 0.7969\n",
      "Epoch 52/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.6263 - accuracy: 0.7670 - val_loss: 0.4924 - val_accuracy: 0.8139\n",
      "Epoch 53/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.6230 - accuracy: 0.7675 - val_loss: 0.4663 - val_accuracy: 0.8329\n",
      "Epoch 54/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.6005 - accuracy: 0.7792 - val_loss: 0.4492 - val_accuracy: 0.8300\n",
      "Epoch 55/70\n",
      "233/233 [==============================] - 15s 63ms/step - loss: 0.5826 - accuracy: 0.7811 - val_loss: 0.4518 - val_accuracy: 0.8342\n",
      "Epoch 56/70\n",
      "233/233 [==============================] - 15s 64ms/step - loss: 0.5856 - accuracy: 0.7764 - val_loss: 0.3771 - val_accuracy: 0.8653\n",
      "Epoch 57/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5589 - accuracy: 0.7940 - val_loss: 0.3786 - val_accuracy: 0.8622\n",
      "Epoch 58/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5532 - accuracy: 0.7988 - val_loss: 0.4224 - val_accuracy: 0.8472\n",
      "Epoch 59/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5389 - accuracy: 0.8013 - val_loss: 0.4306 - val_accuracy: 0.8390\n",
      "Epoch 60/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5299 - accuracy: 0.8056 - val_loss: 0.3615 - val_accuracy: 0.8644\n",
      "Epoch 61/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5179 - accuracy: 0.8074 - val_loss: 0.4165 - val_accuracy: 0.8503\n",
      "Epoch 62/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5113 - accuracy: 0.8139 - val_loss: 0.3583 - val_accuracy: 0.8688\n",
      "Epoch 63/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.5031 - accuracy: 0.8164 - val_loss: 0.3554 - val_accuracy: 0.8677\n",
      "Epoch 64/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.4952 - accuracy: 0.8173 - val_loss: 0.3199 - val_accuracy: 0.8847\n",
      "Epoch 65/70\n",
      "233/233 [==============================] - 15s 62ms/step - loss: 0.4785 - accuracy: 0.8277 - val_loss: 0.3611 - val_accuracy: 0.8655\n",
      "Epoch 66/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.4457 - accuracy: 0.8393 - val_loss: 0.2929 - val_accuracy: 0.8953\n",
      "Epoch 67/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.4599 - accuracy: 0.8260 - val_loss: 0.2800 - val_accuracy: 0.9015\n",
      "Epoch 68/70\n",
      "233/233 [==============================] - 14s 62ms/step - loss: 0.4426 - accuracy: 0.8405 - val_loss: 0.3343 - val_accuracy: 0.8775\n",
      "Epoch 69/70\n",
      "233/233 [==============================] - 15s 64ms/step - loss: 0.4455 - accuracy: 0.8386 - val_loss: 0.3502 - val_accuracy: 0.8697\n",
      "Epoch 70/70\n",
      "233/233 [==============================] - 16s 67ms/step - loss: 0.4294 - accuracy: 0.8452 - val_loss: 0.2500 - val_accuracy: 0.9137\n"
     ]
    }
   ],
   "source": [
    "history = cnn.fit(x = training_set, validation_data = test_set, epochs=70)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PLOTS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "print(history.history.keys())\n",
    "# summarize history for accuracy\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.savefig('D:/Emotion Detector/Accuracy_Graph.png')\n",
    "plt.show()\n",
    "# summarize history for loss\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.savefig('D:/Emotion Detector/Loss_Graph.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "convolutional_neural_network.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}