{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "-UbmiVOrY4nG"
},
"source": [
"#MODEL TRAINING"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "P9mxUB3aY6aB"
},
"outputs": [],
"source": [
"import pickle\n",
"import numpy as np\n",
"import tensorflow as tf\n",
"import keras"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "De0uG17naSnH"
},
"outputs": [],
"source": [
"#Loading the data from pickle file:\n",
"with open('images.p','rb') as f:\n",
" x = pickle.load(f)\n",
"with open('labels.p','rb') as f:\n",
" y = pickle.load(f)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "ocPW78hra7po",
"outputId": "e1deda61-4271-465a-e1f0-5440a7637c74"
},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Images: (16082, 48, 48)\n",
"Labels: (16082,)\n"
]
}
],
"source": [
"print(\"Images: \",x.shape)\n",
"print(\"Labels: \",y.shape)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "r3H-dwisbW1U"
},
"outputs": [],
"source": [
"#Label encoding output\n",
"from sklearn.preprocessing import LabelEncoder\n",
"le = LabelEncoder()\n",
"y = le.fit_transform(y)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "xz9pXqhQb3cz",
"outputId": "0f489cd0-e726-438d-e5fc-932ceb502ada"
},
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"array([0, 0, 0, ..., 4, 4, 4])"
]
},
"metadata": {},
"execution_count": 5
}
],
"source": [
"y"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "l-0sTzUwb4ty"
},
"outputs": [],
"source": [
"# Standardizing the images for CNN Model:\n",
"x = x/255\n",
"x = x.reshape(-1,48,48,1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "pwpKEUdVMx3z"
},
"outputs": [],
"source": [
"# using one-hot encoding to output column\n",
"from keras.utils import to_categorical\n",
"y = to_categorical(y)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "gb_sHsDgNThj"
},
"outputs": [],
"source": [
"# Splitting, training and testing datas\n",
"from sklearn.model_selection import train_test_split\n",
"xtrain, xtest, ytrain, ytest = train_test_split(x,y,test_size=0.1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "6bd6RjU2NzQY",
"outputId": "4de3a15c-e6e0-497d-85f8-25aa758720b9"
},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Xtrain: (14473, 48, 48, 1)\n",
"Xtest: (1609, 48, 48, 1)\n",
"Ytrain: (14473, 5)\n",
"Ytest: (1609, 5)\n"
]
}
],
"source": [
"print(\"Xtrain: \", xtrain.shape)\n",
"print(\"Xtest: \", xtest.shape)\n",
"print(\"Ytrain: \", ytrain.shape)\n",
"print(\"Ytest: \", ytest.shape)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "YTvFdOMFOFw6"
},
"outputs": [],
"source": [
"# CNN Model Training\n",
"from keras.models import Sequential\n",
"from keras.optimizers import Adam\n",
"from keras.layers import Dense, Conv2D, MaxPooling2D, Flatten, Dropout"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "9z_tdS9aPa85",
"outputId": "8a0fc7f0-b8ee-4618-fb36-f9284c02b93c"
},
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"/usr/local/lib/python3.12/dist-packages/keras/src/layers/convolutional/base_conv.py:113: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
" super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
]
}
],
"source": [
"model = Sequential()\n",
"\n",
"# 1st Convolutional Layer:\n",
"model.add(Conv2D(45, (3,3), activation='relu', input_shape=(48,48,1)))\n",
"model.add(MaxPooling2D(pool_size=(2,2)))\n",
"\n",
"# 2nd Convolutional Layer:\n",
"model.add(Conv2D(64, (3,3), activation='relu'))\n",
"model.add(MaxPooling2D(pool_size=(2,2)))\n",
"\n",
"# 3rd Convolutional Layer:\n",
"model.add(Conv2D(128, (3,3), activation='relu'))\n",
"model.add(MaxPooling2D(pool_size=(2,2)))\n",
"\n",
"# Adding flatten layer: converts 2D to 1D\n",
"model.add(Flatten())\n",
"\n",
"# Hidden Layer\n",
"model.add(Dense(256, activation='relu'))\n",
"\n",
"# Output Layer\n",
"model.add(Dense(5, activation='softmax'))\n",
"model.compile(Adam(learning_rate=0.001), loss= 'categorical_crossentropy', metrics=['accuracy'])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 435
},
"id": "tF6iuGvFSHxH",
"outputId": "0dec5c91-6e9d-4b5d-db42-a120e7edd465"
},
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"\u001b[1mModel: \"sequential\"\u001b[0m\n"
],
"text/html": [
"Model: \"sequential\"\n",
"\n"
]
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ conv2d (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m46\u001b[0m, \u001b[38;5;34m46\u001b[0m, \u001b[38;5;34m45\u001b[0m) │ \u001b[38;5;34m450\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m23\u001b[0m, \u001b[38;5;34m23\u001b[0m, \u001b[38;5;34m45\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m21\u001b[0m, \u001b[38;5;34m21\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m25,984\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_1 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m, \u001b[38;5;34m10\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m73,856\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_2 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m4\u001b[0m, \u001b[38;5;34m4\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten (\u001b[38;5;33mFlatten\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2048\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m256\u001b[0m) │ \u001b[38;5;34m524,544\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m5\u001b[0m) │ \u001b[38;5;34m1,285\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
],
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ conv2d (Conv2D) │ (None, 46, 46, 45) │ 450 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d (MaxPooling2D) │ (None, 23, 23, 45) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_1 (Conv2D) │ (None, 21, 21, 64) │ 25,984 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_1 (MaxPooling2D) │ (None, 10, 10, 64) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ conv2d_2 (Conv2D) │ (None, 8, 8, 128) │ 73,856 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ max_pooling2d_2 (MaxPooling2D) │ (None, 4, 4, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ flatten (Flatten) │ (None, 2048) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense (Dense) │ (None, 256) │ 524,544 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (Dense) │ (None, 5) │ 1,285 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
]
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m626,119\u001b[0m (2.39 MB)\n"
],
"text/html": [
" Total params: 626,119 (2.39 MB)\n",
"\n"
]
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m626,119\u001b[0m (2.39 MB)\n"
],
"text/html": [
" Trainable params: 626,119 (2.39 MB)\n",
"\n"
]
},
"metadata": {}
},
{
"output_type": "display_data",
"data": {
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
],
"text/html": [
" Non-trainable params: 0 (0.00 B)\n",
"\n"
]
},
"metadata": {}
}
],
"source": [
"model.summary()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "xk6RG0TCSKLh",
"outputId": "dfb2d5ff-0328-4f68-9408-c9805c4f8f42"
},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Epoch 1/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m65s\u001b[0m 139ms/step - accuracy: 0.2955 - loss: 1.5343 - val_accuracy: 0.4910 - val_loss: 1.2713\n",
"Epoch 2/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m80s\u001b[0m 135ms/step - accuracy: 0.5104 - loss: 1.1988 - val_accuracy: 0.5712 - val_loss: 1.0940\n",
"Epoch 3/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m81s\u001b[0m 133ms/step - accuracy: 0.5894 - loss: 1.0380 - val_accuracy: 0.5855 - val_loss: 1.0413\n",
"Epoch 4/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m60s\u001b[0m 133ms/step - accuracy: 0.6276 - loss: 0.9492 - val_accuracy: 0.6041 - val_loss: 1.0269\n",
"Epoch 5/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m83s\u001b[0m 135ms/step - accuracy: 0.6702 - loss: 0.8558 - val_accuracy: 0.6140 - val_loss: 0.9967\n",
"Epoch 6/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m61s\u001b[0m 135ms/step - accuracy: 0.7116 - loss: 0.7578 - val_accuracy: 0.6109 - val_loss: 1.0013\n",
"Epoch 7/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m60s\u001b[0m 133ms/step - accuracy: 0.7470 - loss: 0.6766 - val_accuracy: 0.6221 - val_loss: 1.0187\n",
"Epoch 8/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m61s\u001b[0m 135ms/step - accuracy: 0.7817 - loss: 0.5893 - val_accuracy: 0.6116 - val_loss: 1.0794\n",
"Epoch 9/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m62s\u001b[0m 137ms/step - accuracy: 0.8095 - loss: 0.5134 - val_accuracy: 0.6271 - val_loss: 1.1883\n",
"Epoch 10/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m61s\u001b[0m 135ms/step - accuracy: 0.8536 - loss: 0.4024 - val_accuracy: 0.6172 - val_loss: 1.2258\n",
"Epoch 11/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m81s\u001b[0m 133ms/step - accuracy: 0.8920 - loss: 0.3159 - val_accuracy: 0.6122 - val_loss: 1.3291\n",
"Epoch 12/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m63s\u001b[0m 138ms/step - accuracy: 0.9156 - loss: 0.2422 - val_accuracy: 0.6022 - val_loss: 1.5368\n",
"Epoch 13/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m62s\u001b[0m 137ms/step - accuracy: 0.9357 - loss: 0.1949 - val_accuracy: 0.5979 - val_loss: 1.7381\n",
"Epoch 14/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m62s\u001b[0m 136ms/step - accuracy: 0.9596 - loss: 0.1256 - val_accuracy: 0.6265 - val_loss: 1.8758\n",
"Epoch 15/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m80s\u001b[0m 133ms/step - accuracy: 0.9655 - loss: 0.1065 - val_accuracy: 0.6097 - val_loss: 2.0449\n",
"Epoch 16/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m85s\u001b[0m 139ms/step - accuracy: 0.9659 - loss: 0.1077 - val_accuracy: 0.6215 - val_loss: 2.2283\n",
"Epoch 17/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m81s\u001b[0m 136ms/step - accuracy: 0.9791 - loss: 0.0665 - val_accuracy: 0.5898 - val_loss: 2.4029\n",
"Epoch 18/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m61s\u001b[0m 135ms/step - accuracy: 0.9751 - loss: 0.0755 - val_accuracy: 0.6277 - val_loss: 2.4831\n",
"Epoch 19/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m83s\u001b[0m 139ms/step - accuracy: 0.9835 - loss: 0.0548 - val_accuracy: 0.6066 - val_loss: 2.5477\n",
"Epoch 20/20\n",
"\u001b[1m453/453\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m80s\u001b[0m 135ms/step - accuracy: 0.9792 - loss: 0.0634 - val_accuracy: 0.6134 - val_loss: 2.7209\n"
]
}
],
"source": [
"history = model.fit(xtrain, ytrain, epochs=20, validation_data=(xtest, ytest))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "dUGGIyMMSxeP",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "1616b37c-016e-42de-8e8f-1aa89983461e"
},
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Model Saved\n"
]
}
],
"source": [
"# to save the model for deployment\n",
"# h5 -> h5 models represent HDF5 format\n",
"# It stores model's optimizers state, model architecture, training configuration, weights\n",
"model.save('emotion_detection_model.h5')\n",
"print(\"Model Saved\")"
]
},
{
"cell_type": "code",
"source": [],
"metadata": {
"id": "IL-mOkAsjJmz"
},
"execution_count": null,
"outputs": []
}
],
"metadata": {
"colab": {
"provenance": []
},
"kernelspec": {
"display_name": "Python 3",
"name": "python3"
},
"language_info": {
"name": "python"
}
},
"nbformat": 4,
"nbformat_minor": 0
}