Face_Mask_Detection.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "YiZChrTbZK_C"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "FoKde6shXG-b"
      },
      "outputs": [],
      "source": [
        "!mkdir -p ~/.kaggle      # setting the directory of kaggle\n",
        "!cp kaggle.json ~/.kaggle/"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ofE6DhikXuQ0"
      },
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PbskT6dlXbRM",
        "outputId": "77e8930b-6258-41c0-fcc0-4c2af248a494"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "face-mask-12k-images-dataset.zip: Skipping, found more recently modified local copy (use --force to force download)\n"
          ]
        }
      ],
      "source": [
        "!kaggle datasets download -d ashishjangra27/face-mask-12k-images-dataset # downloading the dataset from kaggle"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "id": "aMydcToQXdZs"
      },
      "outputs": [],
      "source": [
        "!chmod 600 /root/.kaggle/kaggle.json"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "j8rb_nO0XfmK"
      },
      "outputs": [],
      "source": [
        "import zipfile          # Unzipping the file download from kaggle\n",
        "zip_ref = zipfile.ZipFile('/content/face-mask-12k-images-dataset.zip', 'r')\n",
        "zip_ref.extractall('/content')\n",
        "zip_ref.close()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "ELcQjjSDXiw1"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf    # importing the dependencies\n",
        "from tensorflow import keras\n",
        "from keras import Sequential\n",
        "from keras.layers import Conv2D,Dense,MaxPooling2D,Flatten,BatchNormalization,Dropout\n",
        "from keras.applications.vgg16 import VGG16"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "_7qNspZHZTyN"
      },
      "outputs": [],
      "source": [
        "#Making object of VGG16\n",
        "conv_base=VGG16(\n",
        "    weights=\"imagenet\",\n",
        "    include_top=False,\n",
        "    input_shape=(256,256,3)\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VTomJCaUZcZW",
        "outputId": "d74d3ab2-f989-44ce-8fcf-480507b11b20"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"vgg16\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 256, 256, 3)]     0         \n",
            "                                                                 \n",
            " block1_conv1 (Conv2D)       (None, 256, 256, 64)      1792      \n",
            "                                                                 \n",
            " block1_conv2 (Conv2D)       (None, 256, 256, 64)      36928     \n",
            "                                                                 \n",
            " block1_pool (MaxPooling2D)  (None, 128, 128, 64)      0         \n",
            "                                                                 \n",
            " block2_conv1 (Conv2D)       (None, 128, 128, 128)     73856     \n",
            "                                                                 \n",
            " block2_conv2 (Conv2D)       (None, 128, 128, 128)     147584    \n",
            "                                                                 \n",
            " block2_pool (MaxPooling2D)  (None, 64, 64, 128)       0         \n",
            "                                                                 \n",
            " block3_conv1 (Conv2D)       (None, 64, 64, 256)       295168    \n",
            "                                                                 \n",
            " block3_conv2 (Conv2D)       (None, 64, 64, 256)       590080    \n",
            "                                                                 \n",
            " block3_conv3 (Conv2D)       (None, 64, 64, 256)       590080    \n",
            "                                                                 \n",
            " block3_pool (MaxPooling2D)  (None, 32, 32, 256)       0         \n",
            "                                                                 \n",
            " block4_conv1 (Conv2D)       (None, 32, 32, 512)       1180160   \n",
            "                                                                 \n",
            " block4_conv2 (Conv2D)       (None, 32, 32, 512)       2359808   \n",
            "                                                                 \n",
            " block4_conv3 (Conv2D)       (None, 32, 32, 512)       2359808   \n",
            "                                                                 \n",
            " block4_pool (MaxPooling2D)  (None, 16, 16, 512)       0         \n",
            "                                                                 \n",
            " block5_conv1 (Conv2D)       (None, 16, 16, 512)       2359808   \n",
            "                                                                 \n",
            " block5_conv2 (Conv2D)       (None, 16, 16, 512)       2359808   \n",
            "                                                                 \n",
            " block5_conv3 (Conv2D)       (None, 16, 16, 512)       2359808   \n",
            "                                                                 \n",
            " block5_pool (MaxPooling2D)  (None, 8, 8, 512)         0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 14,714,688\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "#summary of the model\n",
        "conv_base.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "id": "gPRaDMjBZe89"
      },
      "outputs": [],
      "source": [
        "# Freezing the model\n",
        "conv_base.trainable=False"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NJlRF7Y1ZiEC",
        "outputId": "3f326d5c-3101-4aa2-a6ae-9acf5c60a1a4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"vgg16\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 256, 256, 3)]     0         \n",
            "                                                                 \n",
            " block1_conv1 (Conv2D)       (None, 256, 256, 64)      1792      \n",
            "                                                                 \n",
            " block1_conv2 (Conv2D)       (None, 256, 256, 64)      36928     \n",
            "                                                                 \n",
            " block1_pool (MaxPooling2D)  (None, 128, 128, 64)      0         \n",
            "                                                                 \n",
            " block2_conv1 (Conv2D)       (None, 128, 128, 128)     73856     \n",
            "                                                                 \n",
            " block2_conv2 (Conv2D)       (None, 128, 128, 128)     147584    \n",
            "                                                                 \n",
            " block2_pool (MaxPooling2D)  (None, 64, 64, 128)       0         \n",
            "                                                                 \n",
            " block3_conv1 (Conv2D)       (None, 64, 64, 256)       295168    \n",
            "                                                                 \n",
            " block3_conv2 (Conv2D)       (None, 64, 64, 256)       590080    \n",
            "                                                                 \n",
            " block3_conv3 (Conv2D)       (None, 64, 64, 256)       590080    \n",
            "                                                                 \n",
            " block3_pool (MaxPooling2D)  (None, 32, 32, 256)       0         \n",
            "                                                                 \n",
            " block4_conv1 (Conv2D)       (None, 32, 32, 512)       1180160   \n",
            "                                                                 \n",
            " block4_conv2 (Conv2D)       (None, 32, 32, 512)       2359808   \n",
            "                                                                 \n",
            " block4_conv3 (Conv2D)       (None, 32, 32, 512)       2359808   \n",
            "                                                                 \n",
            " block4_pool (MaxPooling2D)  (None, 16, 16, 512)       0         \n",
            "                                                                 \n",
            " block5_conv1 (Conv2D)       (None, 16, 16, 512)       2359808   \n",
            "                                                                 \n",
            " block5_conv2 (Conv2D)       (None, 16, 16, 512)       2359808   \n",
            "                                                                 \n",
            " block5_conv3 (Conv2D)       (None, 16, 16, 512)       2359808   \n",
            "                                                                 \n",
            " block5_pool (MaxPooling2D)  (None, 8, 8, 512)         0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 14,714,688\n",
            "Trainable params: 0\n",
            "Non-trainable params: 14,714,688\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "conv_base.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "id": "x8V92CarZkGu"
      },
      "outputs": [],
      "source": [
        "# Adding the ANN part to the end of VGG16\n",
        "model=Sequential()\n",
        "\n",
        "model.add(conv_base)\n",
        "model.add(Flatten())\n",
        "model.add(Dense(256,activation=\"relu\"))\n",
        "model.add(Dropout(0.3))\n",
        "model.add(Dense(1,activation=\"sigmoid\"))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4ZkE0fGsZnIU",
        "outputId": "29ecb380-3ba8-43cd-d5a7-012ac5869321"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " vgg16 (Functional)          (None, 8, 8, 512)         14714688  \n",
            "                                                                 \n",
            " flatten_1 (Flatten)         (None, 32768)             0         \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 256)               8388864   \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 1)                 257       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 23,103,809\n",
            "Trainable params: 8,389,121\n",
            "Non-trainable params: 14,714,688\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 38,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ONp1dpC9ZpIw",
        "outputId": "2446a5d7-bcbc-4fbe-904d-157386515811"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 10000 files belonging to 2 classes.\n",
            "Found 992 files belonging to 2 classes.\n"
          ]
        }
      ],
      "source": [
        "#Making Train and Test data\n",
        "\n",
        "train_data = keras.utils.image_dataset_from_directory(\n",
        "    directory=\"/content/Face Mask Dataset/Train\",\n",
        "    labels=\"inferred\",\n",
        "    label_mode=\"int\",\n",
        "    batch_size=32,\n",
        "    image_size=(256,256)\n",
        ")\n",
        "test_data=keras.utils.image_dataset_from_directory(\n",
        "    directory=\"/content/Face Mask Dataset/Test\",\n",
        "    labels=\"inferred\",\n",
        "    label_mode=\"int\",\n",
        "    batch_size=32,\n",
        "    image_size=(256,256)\n",
        ")\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 39,
      "metadata": {
        "id": "5l6WscI_aN1g"
      },
      "outputs": [],
      "source": [
        "# Standardizing the image size\n",
        "def process(image,label):\n",
        "  image=tf.cast(image/255,tf.float32)\n",
        "\n",
        "  return image,label\n",
        "\n",
        "train_data=train_data.map(process)\n",
        "#test_data=test_data.map(process)"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "vaY9W99AoF_r"
      },
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "id": "C0F5JAtTaSmO"
      },
      "outputs": [],
      "source": [
        "# compiling the model\n",
        "model.compile(optimizer=\"adam\",loss=\"binary_crossentropy\",metrics=[\"accuracy\"])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 43,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ioGl-T0waV6z",
        "outputId": "631201a6-563e-4eb5-ac98-de5b5a36d6b2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "313/313 [==============================] - 69s 218ms/step - loss: 0.0280 - accuracy: 0.9897 - val_loss: 0.4838 - val_accuracy: 0.9929\n",
            "Epoch 2/5\n",
            "313/313 [==============================] - 61s 192ms/step - loss: 0.0152 - accuracy: 0.9941 - val_loss: 0.8778 - val_accuracy: 0.9899\n",
            "Epoch 3/5\n",
            "313/313 [==============================] - 61s 192ms/step - loss: 0.0187 - accuracy: 0.9924 - val_loss: 1.0321 - val_accuracy: 0.9889\n",
            "Epoch 4/5\n",
            "313/313 [==============================] - 64s 203ms/step - loss: 0.0128 - accuracy: 0.9953 - val_loss: 0.8287 - val_accuracy: 0.9909\n",
            "Epoch 5/5\n",
            "313/313 [==============================] - 60s 190ms/step - loss: 0.0085 - accuracy: 0.9970 - val_loss: 0.5315 - val_accuracy: 0.9899\n"
          ]
        }
      ],
      "source": [
        "# train the model\n",
        "history=model.fit(train_data,epochs=5,validation_data = (test_data))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 49,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 152
        },
        "id": "jn0gDVU1aZCm",
        "outputId": "cb12c183-f03d-4f71-8028-9714835b1539"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=100x100>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 29ms/step\n",
            "With Mask\n"
          ]
        }
      ],
      "source": [
        "# testing the model with some random images\n",
        "import cv2\n",
        "from google.colab.patches import cv2_imshow\n",
        "\n",
        "img = cv2.imread(\"/content/Face Mask Dataset/Validation/WithMask/1002.png\")\n",
        "\n",
        "cv2_imshow(img)\n",
        "\n",
        "def prediction(img):\n",
        "  img_sh=cv2.resize(img,(256,256))\n",
        "  test_img=img_sh.reshape((1,256,256,3))\n",
        "\n",
        "  output=model.predict(test_img)\n",
        "  ot=int(output[0][0])\n",
        "\n",
        "  if ot == 0:\n",
        "\n",
        "    print(\"With Mask\")\n",
        "\n",
        "  else:\n",
        "    print(\"Without Mask\")\n",
        "\n",
        "prediction(img)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 50,
      "metadata": {
        "id": "sgenhBJJd7Fs"
      },
      "outputs": [],
      "source": [
        "import pickle\n",
        "\n",
        "pickle.dump(model,open('facemask detection model.pkl','wb'))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 51,
      "metadata": {
        "id": "TffcY0qFfOYc"
      },
      "outputs": [],
      "source": [
        "\n",
        "tf.keras.models.save_model(model,\"my_model.hdf5\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nw0KlHoRd47g"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}