Training_Emotion.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Training_Emotion.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "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/PranavKathar/DeepVision-Emotion-and-Gesture-Detection/blob/master/Training_Emotion.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1IywQ_KPHCjl"
      },
      "outputs": [],
      "source": [
        "import keras\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, Dropout, BatchNormalization, Dense, Flatten, Activation\n",
        "from tensorflow.keras.models import Sequential, Model\n",
        "from tensorflow.keras.regularizers import l2\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "import numpy as np\n",
        "import pydot_ng\n",
        "import graphviz\n",
        "import os\n",
        "\n",
        "num_classes = 6\n",
        "\n",
        "Img_Height = 48\n",
        "Img_width = 48\n",
        "\n",
        "batch_size = 32\n",
        "train_directory = \"drive/MyDrive/emotion_2/train\"\n",
        "train_dir = os.listdir(train_directory)\n",
        "validation_directory = \"drive/MyDrive/emotion_2/validation\"\n",
        "validation_dir = os.listdir(validation_directory)\n",
        "\n",
        "\n",
        "train_datagen = ImageDataGenerator(rescale=1./255,\n",
        "                                   rotation_range=60,\n",
        "                                   shear_range=0.5,\n",
        "                                   zoom_range=0.5,\n",
        "                                   width_shift_range=0.5,\n",
        "                                   height_shift_range=0.5,\n",
        "                                   horizontal_flip=True,\n",
        "                                   fill_mode='nearest')\n",
        "\n",
        "validation_datagen = ImageDataGenerator(rescale=1./255)\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(train_directory,\n",
        "                                                    color_mode='grayscale',\n",
        "                                                    target_size=(Img_Height, Img_width),\n",
        "                                                    batch_size=batch_size,\n",
        "                                                    class_mode='categorical',\n",
        "                                                    shuffle=True)\n",
        "\n",
        "validation_generator = validation_datagen.flow_from_directory(validation_directory,\n",
        "                                                              color_mode='grayscale',\n",
        "                                                              target_size=(Img_Height, Img_width),\n",
        "                                                              batch_size=batch_size,\n",
        "                                                              class_mode='categorical',\n",
        "                                                              shuffle=True)\n",
        "\n",
        "\n",
        "model = Sequential()\n",
        "\n",
        "# Block-1: The First Convolutional Block\n",
        "\n",
        "model.add(Conv2D(filters=32, kernel_size=(3,3), padding='same', \n",
        "                 kernel_initializer='he_normal',\n",
        "                 activation=\"elu\", \n",
        "                 input_shape=(Img_Height, Img_width, 1), \n",
        "                 name=\"Conv1\"))\n",
        "\n",
        "model.add(BatchNormalization(name=\"Batch_Norm1\"))\n",
        "\n",
        "model.add(Conv2D(filters=32, kernel_size=(3,3), padding='same', \n",
        "                 kernel_initializer='he_normal', \n",
        "                 activation=\"elu\", name=\"Conv2\"))\n",
        "\n",
        "model.add(BatchNormalization(name=\"Batch_Norm2\"))\n",
        "model.add(MaxPooling2D(pool_size=(2,2), name=\"Maxpool1\"))\n",
        "model.add(Dropout(0.2, name=\"Dropout1\"))\n",
        "\n",
        "# Block-2: The Second Convolutional Block\n",
        "\n",
        "model.add(Conv2D(filters=64, kernel_size=(3,3), padding='same', \n",
        "                 kernel_initializer='he_normal',\n",
        "                 activation=\"elu\", name=\"Conv3\"))\n",
        "\n",
        "model.add(BatchNormalization(name=\"Batch_Norm3\"))\n",
        "\n",
        "model.add(Conv2D(filters=64,kernel_size=(3,3),padding='same',\n",
        "                 kernel_initializer='he_normal', \n",
        "                 activation=\"elu\", name=\"Conv4\"))\n",
        "\n",
        "model.add(BatchNormalization(name=\"Batch_Norm4\"))\n",
        "model.add(MaxPooling2D(pool_size=(2,2), name=\"Maxpool2\"))\n",
        "model.add(Dropout(0.2, name=\"Dropout2\"))\n",
        "\n",
        "# Block-3: The Third Convolutional Block\n",
        "\n",
        "model.add(Conv2D(filters=128, kernel_size=(3,3), padding='same', \n",
        "                 kernel_initializer='he_normal', \n",
        "                 activation=\"elu\", name=\"Conv5\"))\n",
        "\n",
        "model.add(BatchNormalization(name=\"Batch_Norm5\"))\n",
        "\n",
        "model.add(Conv2D(filters=128, kernel_size=(3,3), padding='same', \n",
        "                 kernel_initializer='he_normal',\n",
        "                 activation=\"elu\", name=\"Conv6\"))\n",
        "\n",
        "model.add(BatchNormalization(name=\"Batch_Norm6\"))\n",
        "model.add(MaxPooling2D(pool_size=(2,2), name=\"Maxpool3\"))\n",
        "model.add(Dropout(0.2, name=\"Dropout3\"))\n",
        "\n",
        "# Block-4: The Fully Connected Block\n",
        "\n",
        "model.add(Flatten(name=\"Flatten\"))\n",
        "model.add(Dense(64, activation=\"elu\", kernel_initializer='he_normal', name=\"Dense\"))\n",
        "model.add(BatchNormalization(name=\"Batch_Norm7\"))\n",
        "model.add(Dropout(0.5, name=\"Dropout4\"))\n",
        "\n",
        "# Block-5: The Output Block\n",
        "\n",
        "model.add(Dense(num_classes, activation=\"softmax\", kernel_initializer='he_normal', name = \"Output\"))\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "from tensorflow.keras.callbacks import ModelCheckpoint\n",
        "from tensorflow.keras.callbacks import ReduceLROnPlateau\n",
        "from tensorflow.keras.callbacks import TensorBoard\n",
        "\n",
        "checkpoint = ModelCheckpoint(\"drive/MyDrive/emotion_2/saved_model/emotions_part1.h5\", monitor='accuracy', verbose=1,\n",
        "                              save_best_only=True, mode='auto', period=1)\n",
        "\n",
        "reduce = ReduceLROnPlateau(monitor='accuracy', factor=0.2, patience=10, \n",
        "                           min_lr=0.0001, verbose = 1)\n",
        "\n",
        "logdir='logs'\n",
        "tensorboard_Visualization = TensorBoard(log_dir=logdir, histogram_freq=False)\n",
        "\n",
        "\n",
        "\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "              optimizer = Adam(lr = 0.001),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "\n",
        "train_samples = 28353\n",
        "validation_samples = 3534\n",
        "epochs = 150\n",
        "batch_size = 64\n",
        "\n",
        "model.fit(train_generator,\n",
        "          steps_per_epoch = train_samples//batch_size,\n",
        "          epochs = epochs,\n",
        "          callbacks = [checkpoint, reduce, tensorboard_Visualization],\n",
        "          validation_data = validation_generator,\n",
        "          validation_steps = validation_samples//batch_size,\n",
        "          shuffle=True)\n",
        "\n",
        "\n",
        "model.save('drive/MyDrive/emotion_2/saved_model/em_part1.h5')\n"
      ]
    }
  ]
}