Mask Detection with Deep Learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras.layers import Flatten, Dense\n",
    "import cv2\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_datagen = ImageDataGenerator(rescale=1./255,\n",
    "                                  shear_range=0.2,\n",
    "                                  zoom_range=0.2)\n",
    "test_datagen = ImageDataGenerator(rescale=1./255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 958 images belonging to 2 classes.\n",
      "Found 406 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "train_set = train_datagen.flow_from_directory('C:\\\\Users\\\\Arpitha Ananth\\\\Documents\\\\Programming in Data Science\\\\Practice Py\\\\Knowledge Base\\\\Face Recognition\\\\Mask Recogonition\\\\data\\\\Train Data',\n",
    "                                               target_size=(64, 64),class_mode='binary')\n",
    "test_set = test_datagen.flow_from_directory('C:\\\\Users\\\\Arpitha Ananth\\\\Documents\\\\Programming in Data Science\\\\Practice Py\\\\Knowledge Base\\\\Face Recognition\\\\Mask Recogonition\\\\data\\\\Test Data',\n",
    "                                            target_size=(64, 64),class_mode='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Arpitha Ananth\\Anaconda3\\envs\\TensorFlow1\\lib\\site-packages\\tensorflow_core\\python\\ops\\resource_variable_ops.py:1630: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "If using Keras pass *_constraint arguments to layers.\n",
      "WARNING:tensorflow:From C:\\Users\\Arpitha Ananth\\Anaconda3\\envs\\TensorFlow1\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:4070: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
      "\n",
      "WARNING:tensorflow:From C:\\Users\\Arpitha Ananth\\Anaconda3\\envs\\TensorFlow1\\lib\\site-packages\\tensorflow_core\\python\\ops\\nn_impl.py:183: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
     ]
    }
   ],
   "source": [
    "classifier = Sequential()\n",
    "classifier.add(Conv2D(64,(3,3),input_shape = (64,64,3), activation = 'relu'))\n",
    "classifier.add(MaxPooling2D(pool_size = (2,2)))\n",
    "\n",
    "classifier.add(Conv2D(32,(3,3), activation = 'relu'))\n",
    "classifier.add(MaxPooling2D(pool_size = (2,2)))\n",
    "\n",
    "classifier.add(Flatten())\n",
    "classifier.add(Dense(units = 1, activation = 'sigmoid'))\n",
    "\n",
    "classifier.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 62, 62, 64)        1792      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 31, 31, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 29, 29, 32)        18464     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 6272)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 1)                 6273      \n",
      "=================================================================\n",
      "Total params: 26,529\n",
      "Trainable params: 26,529\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "classifier.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Arpitha Ananth\\Anaconda3\\envs\\TensorFlow1\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n",
      "Epoch 1/10\n",
      "30/30 [==============================] - 4s 136ms/step - loss: 0.5824 - accuracy: 0.7223 - val_loss: 0.3948 - val_accuracy: 0.8276\n",
      "Epoch 2/10\n",
      "30/30 [==============================] - 4s 129ms/step - loss: 0.2303 - accuracy: 0.9228 - val_loss: 0.3020 - val_accuracy: 0.9163\n",
      "Epoch 3/10\n",
      "30/30 [==============================] - 4s 118ms/step - loss: 0.1313 - accuracy: 0.9541 - val_loss: 0.0305 - val_accuracy: 0.9458\n",
      "Epoch 4/10\n",
      "30/30 [==============================] - 4s 131ms/step - loss: 0.1041 - accuracy: 0.9645 - val_loss: 0.0855 - val_accuracy: 0.9507\n",
      "Epoch 5/10\n",
      "30/30 [==============================] - 3s 111ms/step - loss: 0.0877 - accuracy: 0.9718 - val_loss: 0.0494 - val_accuracy: 0.9655\n",
      "Epoch 6/10\n",
      "30/30 [==============================] - 4s 138ms/step - loss: 0.0771 - accuracy: 0.9770 - val_loss: 0.1028 - val_accuracy: 0.9754\n",
      "Epoch 7/10\n",
      "30/30 [==============================] - 4s 118ms/step - loss: 0.0747 - accuracy: 0.9802 - val_loss: 0.0243 - val_accuracy: 0.9754\n",
      "Epoch 8/10\n",
      "30/30 [==============================] - 5s 159ms/step - loss: 0.0631 - accuracy: 0.9854 - val_loss: 0.1077 - val_accuracy: 0.9778\n",
      "Epoch 9/10\n",
      "30/30 [==============================] - 4s 138ms/step - loss: 0.0702 - accuracy: 0.9823 - val_loss: 0.0700 - val_accuracy: 0.9581\n",
      "Epoch 10/10\n",
      "30/30 [==============================] - 3s 112ms/step - loss: 0.0545 - accuracy: 0.9833 - val_loss: 0.1242 - val_accuracy: 0.9803\n"
     ]
    }
   ],
   "source": [
    "history= classifier.fit_generator(\n",
    "        train_set,\n",
    "        epochs=10,\n",
    "        steps_per_epoch=30,\n",
    "        shuffle=True,\n",
    "        validation_data=test_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "from matplotlib import pyplot\n",
    "pyplot.plot(history.history['loss'], label='train')\n",
    "pyplot.plot(history.history['val_loss'], label='test')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "from matplotlib import pyplot\n",
    "pyplot.plot(history.history['accuracy'], label='train')\n",
    "pyplot.plot(history.history['val_accuracy'], label='test')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "face_cascade = cv2.CascadeClassifier(r'C:\\Users\\Arpitha Ananth\\Documents\\Programming in Data Science\\Practice Py\\Knowledge Base\\Computer Vision\\Module_1_Face_Recognition\\haarcascade_frontalface_default.xml')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "video_capture = cv2.VideoCapture(0)\n",
    "fourcc = cv2.VideoWriter_fourcc(*'DIVX') \n",
    "out = cv2.VideoWriter('check1.mp4', fourcc, 15.0, (640, 480)) \n",
    "\n",
    "while True:\n",
    "    ret,frame =video_capture.read() \n",
    "    res_frame = cv2.resize(frame, (frame.shape[1] // 4, frame.shape[0] //4))\n",
    "    faces = face_cascade.detectMultiScale(res_frame)\n",
    "    for f in faces:\n",
    "        (x, y, w, h) = [v * 4 for v in f] \n",
    "        face_img = frame[y:y+h, x:x+w]\n",
    "        resized=cv2.resize(face_img,(64,64))\n",
    "        normalized=resized/255.0\n",
    "        reshaped=np.reshape(normalized,(1,64,64,3))\n",
    "        reshaped = np.vstack([reshaped])\n",
    "        result=classifier.predict(reshaped)\n",
    "        if result >= 0.6:\n",
    "            prediction = 'Mask'\n",
    "        else:\n",
    "            prediction = 'No Mask'\n",
    "        cv2.rectangle(frame,(x,y),(x+w,y+h),(0,255,255),2)\n",
    "        cv2.putText(frame, prediction, (x, y-10),cv2.FONT_HERSHEY_SIMPLEX,0.8,(255,255,255),2)\n",
    "    \n",
    "    out.write(frame)\n",
    "    cv2.imshow('Video',frame)\n",
    "    key = cv2.waitKey(10)\n",
    "    if key == 27: #The Esc key\n",
    "        break\n",
    "video_capture.release()\n",
    "cv2.destroyAllWindows()"
   ]
  }
 ],
 "metadata": {
  "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}