train_mask_detector.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[INFO] loading images...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\tiwar\\anaconda3\\lib\\site-packages\\PIL\\Image.py:960: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:`input_shape` is undefined or non-square, or `rows` is not in [96, 128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet_v2/mobilenet_v2_weights_tf_dim_ordering_tf_kernels_1.0_224_no_top.h5\n",
      "9412608/9406464 [==============================] - 1s 0us/step\n",
      "[INFO] compiling model...\n",
      "[INFO] training head...\n",
      "Epoch 1/20\n",
      "95/95 [==============================] - 85s 894ms/step - loss: 0.4410 - accuracy: 0.7980 - val_loss: 0.1002 - val_accuracy: 0.9844\n",
      "Epoch 2/20\n",
      "95/95 [==============================] - 74s 781ms/step - loss: 0.1453 - accuracy: 0.9463 - val_loss: 0.0613 - val_accuracy: 0.9883\n",
      "Epoch 3/20\n",
      "95/95 [==============================] - 75s 790ms/step - loss: 0.0996 - accuracy: 0.9654 - val_loss: 0.0475 - val_accuracy: 0.9883\n",
      "Epoch 4/20\n",
      "95/95 [==============================] - 73s 774ms/step - loss: 0.0647 - accuracy: 0.9806 - val_loss: 0.0451 - val_accuracy: 0.9870\n",
      "Epoch 5/20\n",
      "95/95 [==============================] - 75s 791ms/step - loss: 0.0602 - accuracy: 0.9779 - val_loss: 0.0392 - val_accuracy: 0.9883\n",
      "Epoch 6/20\n",
      "95/95 [==============================] - 78s 825ms/step - loss: 0.0540 - accuracy: 0.9796 - val_loss: 0.0397 - val_accuracy: 0.9896\n",
      "Epoch 7/20\n",
      "95/95 [==============================] - 80s 839ms/step - loss: 0.0518 - accuracy: 0.9835 - val_loss: 0.0382 - val_accuracy: 0.9896\n",
      "Epoch 8/20\n",
      "95/95 [==============================] - 76s 804ms/step - loss: 0.0473 - accuracy: 0.9832 - val_loss: 0.0331 - val_accuracy: 0.9909\n",
      "Epoch 9/20\n",
      "95/95 [==============================] - 76s 800ms/step - loss: 0.0395 - accuracy: 0.9871 - val_loss: 0.0302 - val_accuracy: 0.9909\n",
      "Epoch 10/20\n",
      "95/95 [==============================] - 76s 801ms/step - loss: 0.0374 - accuracy: 0.9888 - val_loss: 0.0313 - val_accuracy: 0.9909\n",
      "Epoch 11/20\n",
      "95/95 [==============================] - 79s 828ms/step - loss: 0.0324 - accuracy: 0.9875 - val_loss: 0.0311 - val_accuracy: 0.9909\n",
      "Epoch 12/20\n",
      "95/95 [==============================] - 76s 800ms/step - loss: 0.0290 - accuracy: 0.9908 - val_loss: 0.0323 - val_accuracy: 0.9909\n",
      "Epoch 13/20\n",
      "95/95 [==============================] - 76s 800ms/step - loss: 0.0321 - accuracy: 0.9881 - val_loss: 0.0282 - val_accuracy: 0.9922\n",
      "Epoch 14/20\n",
      "95/95 [==============================] - 76s 797ms/step - loss: 0.0298 - accuracy: 0.9888 - val_loss: 0.0280 - val_accuracy: 0.9922\n",
      "Epoch 15/20\n",
      "95/95 [==============================] - 77s 812ms/step - loss: 0.0345 - accuracy: 0.9895 - val_loss: 0.0328 - val_accuracy: 0.9909\n",
      "Epoch 16/20\n",
      "95/95 [==============================] - 78s 821ms/step - loss: 0.0215 - accuracy: 0.9924 - val_loss: 0.0277 - val_accuracy: 0.9935\n",
      "Epoch 17/20\n",
      "95/95 [==============================] - 76s 800ms/step - loss: 0.0268 - accuracy: 0.9921 - val_loss: 0.0319 - val_accuracy: 0.9909\n",
      "Epoch 18/20\n",
      "95/95 [==============================] - 76s 800ms/step - loss: 0.0214 - accuracy: 0.9937 - val_loss: 0.0282 - val_accuracy: 0.9935\n",
      "Epoch 19/20\n",
      "95/95 [==============================] - 77s 815ms/step - loss: 0.0305 - accuracy: 0.9924 - val_loss: 0.0405 - val_accuracy: 0.9870\n",
      "Epoch 20/20\n",
      "95/95 [==============================] - 76s 801ms/step - loss: 0.0249 - accuracy: 0.9904 - val_loss: 0.0268 - val_accuracy: 0.9909\n",
      "[INFO] evaluating network...\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "   with_mask       0.99      0.99      0.99       383\n",
      "without_mask       0.99      0.99      0.99       384\n",
      "\n",
      "    accuracy                           0.99       767\n",
      "   macro avg       0.99      0.99      0.99       767\n",
      "weighted avg       0.99      0.99      0.99       767\n",
      "\n",
      "[INFO] saving mask detector model...\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import MobileNetV2\n",
    "from tensorflow.keras.layers import AveragePooling2D\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from tensorflow.keras.layers import Flatten\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.layers import Input\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.applications.mobilenet_v2 import preprocess_input\n",
    "from tensorflow.keras.preprocessing.image import img_to_array\n",
    "from tensorflow.keras.preprocessing.image import load_img\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from sklearn.preprocessing import LabelBinarizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report\n",
    "from imutils import paths\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "# initialize the initial learning rate, number of epochs to train for,\n",
    "# and batch size\n",
    "INIT_LR = 1e-4\n",
    "EPOCHS = 20\n",
    "BS = 32\n",
    "\n",
    "DIRECTORY = r\"C:\\Users\\USERNAME\\Face-Mask-Detection-master (1)\\Face-Mask-Detection-master\\dataset\"\n",
    "CATEGORIES = [\"with_mask\", \"without_mask\"]\n",
    "\n",
    "# grab the list of images in our dataset directory, then initialize\n",
    "# the list of data (i.e., images) and class images\n",
    "print(\"[INFO] loading images...\")\n",
    "\n",
    "data = []\n",
    "labels = []\n",
    "\n",
    "for category in CATEGORIES:\n",
    "    path = os.path.join(DIRECTORY, category)\n",
    "    for img in os.listdir(path):\n",
    "    \timg_path = os.path.join(path, img)\n",
    "    \timage = load_img(img_path, target_size=(224, 224))\n",
    "    \timage = img_to_array(image)\n",
    "    \timage = preprocess_input(image)\n",
    "\n",
    "    \tdata.append(image)\n",
    "    \tlabels.append(category)\n",
    "\n",
    "# perform one-hot encoding on the labels\n",
    "lb = LabelBinarizer()\n",
    "labels = lb.fit_transform(labels)\n",
    "labels = to_categorical(labels)\n",
    "\n",
    "data = np.array(data, dtype=\"float32\")\n",
    "labels = np.array(labels)\n",
    "\n",
    "(trainX, testX, trainY, testY) = train_test_split(data, labels,\n",
    "\ttest_size=0.20, stratify=labels, random_state=42)\n",
    "\n",
    "# construct the training image generator for data augmentation\n",
    "aug = ImageDataGenerator(\n",
    "\trotation_range=20,\n",
    "\tzoom_range=0.15,\n",
    "\twidth_shift_range=0.2,\n",
    "\theight_shift_range=0.2,\n",
    "\tshear_range=0.15,\n",
    "\thorizontal_flip=True,\n",
    "\tfill_mode=\"nearest\")\n",
    "\n",
    "# load the MobileNetV2 network, ensuring the head FC layer sets are\n",
    "# left off\n",
    "baseModel = MobileNetV2(weights=\"imagenet\", include_top=False,\n",
    "\tinput_tensor=Input(shape=(224, 224, 3)))\n",
    "\n",
    "# construct the head of the model that will be placed on top of the\n",
    "# the base model\n",
    "headModel = baseModel.output\n",
    "headModel = AveragePooling2D(pool_size=(7, 7))(headModel)\n",
    "headModel = Flatten(name=\"flatten\")(headModel)\n",
    "headModel = Dense(128, activation=\"relu\")(headModel)\n",
    "headModel = Dropout(0.5)(headModel)\n",
    "headModel = Dense(2, activation=\"softmax\")(headModel)\n",
    "\n",
    "# place the head FC model on top of the base model (this will become\n",
    "# the actual model we will train)\n",
    "model = Model(inputs=baseModel.input, outputs=headModel)\n",
    "\n",
    "# loop over all layers in the base model and freeze them so they will\n",
    "# *not* be updated during the first training process\n",
    "for layer in baseModel.layers:\n",
    "\tlayer.trainable = False\n",
    "\n",
    "# compile our model\n",
    "print(\"[INFO] compiling model...\")\n",
    "opt = Adam(lr=INIT_LR, decay=INIT_LR / EPOCHS)\n",
    "model.compile(loss=\"binary_crossentropy\", optimizer=opt,\n",
    "\tmetrics=[\"accuracy\"])\n",
    "\n",
    "# train the head of the network\n",
    "print(\"[INFO] training head...\")\n",
    "H = model.fit(\n",
    "\taug.flow(trainX, trainY, batch_size=BS),\n",
    "\tsteps_per_epoch=len(trainX) // BS,\n",
    "\tvalidation_data=(testX, testY),\n",
    "\tvalidation_steps=len(testX) // BS,\n",
    "\tepochs=EPOCHS)\n",
    "\n",
    "# make predictions on the testing set\n",
    "print(\"[INFO] evaluating network...\")\n",
    "predIdxs = model.predict(testX, batch_size=BS)\n",
    "\n",
    "# for each image in the testing set we need to find the index of the\n",
    "# label with corresponding largest predicted probability\n",
    "predIdxs = np.argmax(predIdxs, axis=1)\n",
    "\n",
    "# show a nicely formatted classification report\n",
    "print(classification_report(testY.argmax(axis=1), predIdxs,\n",
    "\ttarget_names=lb.classes_))\n",
    "\n",
    "# serialize the model to disk\n",
    "print(\"[INFO] saving mask detector model...\")\n",
    "model.save(\"mask_detector.model\", save_format=\"h5\")\n",
    "\n",
    "# plot the training loss and accuracy\n",
    "N = EPOCHS\n",
    "plt.style.use(\"ggplot\")\n",
    "plt.figure()\n",
    "plt.plot(np.arange(0, N), H.history[\"loss\"], label=\"train_loss\")\n",
    "plt.plot(np.arange(0, N), H.history[\"val_loss\"], label=\"val_loss\")\n",
    "plt.plot(np.arange(0, N), H.history[\"accuracy\"], label=\"train_acc\")\n",
    "plt.plot(np.arange(0, N), H.history[\"val_accuracy\"], label=\"val_acc\")\n",
    "plt.title(\"Training Loss and Accuracy\")\n",
    "plt.xlabel(\"Epoch #\")\n",
    "plt.ylabel(\"Loss/Accuracy\")\n",
    "plt.legend(loc=\"lower left\")\n",
    "plt.savefig(\"plot.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}