3_Face_mask_detection_corrigé.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "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/hocine-drioueche/OpenclassroomsProject/blob/main/3_Face_mask_detection_corrig%C3%A9.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y7KtILhji3MF"
      },
      "source": [
        "<center><img src=\"https://assets-datascientest.s3-eu-west-1.amazonaws.com/train/logo_datascientest.png\" style=\"height:150px;center\"></center>\n",
        "\n",
        "<hr style=\"border-width:2px;border-color:#75DFC1\">\n",
        "<center><h1>Deep-Learning - Modules complémentaires</h1></center>\n",
        "<center><h2>Détection de masque</h2></center>\n",
        "<hr style=\"border-width:2px;border-color:#75DFC1\">\n",
        "\n",
        "\n",
        "Le but de cet exercice est de s'exercer au transfer learning.\n",
        "\n",
        " Nous allons travailler avec la base de données **```Face Mask Dataset```**. Cette base contient des images de personnes avec et sans masque de protection contre la Covid-19.\n",
        "\n",
        " Notez que les images ne sont pas stockées dans un DataFrame (comme lors de l'exercice sur l'âge et le genre). Nous avons directement accès aux liens des images. Nous allons pouvoir effectuer le preprocessing des images très rapidement avec des générateurs en utilisant la méthode ```flow_from_directory```.\n",
        "\n",
        "> La structure de l'exercice est la suivante :\n",
        ">> I - [Traitement du dataset](#preparation)\n",
        ">>> A - [Affichage des images](#affichage)\n",
        ">>>\n",
        ">>>\n",
        ">>> B - [Générateurs](#generateurs)\n",
        ">>\n",
        ">>\n",
        ">> II - [Classification](#classif)\n",
        ">>\n",
        ">>\n",
        ">> III - [Test en temps réel](#test)\n",
        "\n",
        "- Exécuter les cellules ci-dessous pour importer le dataset et les modules nécessaires à l'exercice."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cWRRr-zCHHvw",
        "outputId": "669cb7ad-9d3e-428c-b3c2-bee9e78e6bfd"
      },
      "source": [
        "!wget https://assets-datascientest.s3-eu-west-1.amazonaws.com/datasets/Face+Mask+Dataset.zip\n",
        "!unzip Face+Mask+Dataset.zip"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2022-01-03 14:03:33--  https://assets-datascientest.s3-eu-west-1.amazonaws.com/datasets/Face+Mask+Dataset.zip\n",
            "Resolving assets-datascientest.s3-eu-west-1.amazonaws.com (assets-datascientest.s3-eu-west-1.amazonaws.com)... 52.218.101.208\n",
            "Connecting to assets-datascientest.s3-eu-west-1.amazonaws.com (assets-datascientest.s3-eu-west-1.amazonaws.com)|52.218.101.208|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 19487225 (19M) [application/zip]\n",
            "Saving to: ‘Face+Mask+Dataset.zip’\n",
            "\n",
            "Face+Mask+Dataset.z 100%[===================>]  18.58M  9.99MB/s    in 1.9s    \n",
            "\n",
            "2022-01-03 14:03:36 (9.99 MB/s) - ‘Face+Mask+Dataset.zip’ saved [19487225/19487225]\n",
            "\n",
            "Archive:  Face+Mask+Dataset.zip\n",
            "  inflating: Face Mask Dataset/Train/WithMask/532.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/978.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1100.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/963.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/915.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/929.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/502.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/630.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/251.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/95.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/618.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/102.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/601.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/855.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/121.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/129.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1117.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1077.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/89.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/491.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/257.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/142.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/113.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/690.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/158.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/170.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/766.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1059.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/965.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/120.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/971.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/162.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1005.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/366.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/961.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/131.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/775.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/359.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/804.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/119.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/123.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1038.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/790.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/948.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/958.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/81.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/148.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/109.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/126.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/810.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/115.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/970.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/139.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/164.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/81.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1412.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/964.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1199.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/974.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/94.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/80.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/784.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/88.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/828.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/112.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/800.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/96.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/137.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/960.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/94.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/99.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/814.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/794.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1200.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1228.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/89.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/98.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/82.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/503.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/80.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1638.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/540.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/545.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1562.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/743.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/477.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/480.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2651.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/489.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/95.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3358.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/496.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2735.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3166.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2720.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/499.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/475.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/476.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/490.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/467.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/471.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/504.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/491.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/469.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/502.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2682.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2443.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2872.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3588.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1837.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/743.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/838.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/958.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/753.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2523.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1416.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2719.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3407.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2686.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2876.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3615.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3600.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1370.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3563.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/83.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2457.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1823.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3229.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2899.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/394.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2490.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2536.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/226.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1765.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/425.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/800.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1228.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1566.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3198.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2709.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1176.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/357.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1589.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1017.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2527.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1412.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/757.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1176.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/232.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/627.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1600.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/425.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/633.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/555.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/569.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1002.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1628.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/117.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1764.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/130.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/154.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/998.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1758.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/971.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/91.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/92.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/807.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1348.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/551.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1374.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1016.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1204.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/409.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/431.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/747.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1402.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1358.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1576.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/419.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/138.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/754.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/435.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/133.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/90.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/769.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1566.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/86.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1406.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/969.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/780.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/770.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/79.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/304.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/395.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/768.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/255.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/742.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/740.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/53.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/350.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/197.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/181.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1105.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/608.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/922.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/242.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/548.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/171.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1084.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/683.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/319.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/635.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/422.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/696.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/643.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1024.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/556.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1080.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/542.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/996.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/478.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/474.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/737.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/501.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/492.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/486.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/483.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/473.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1037.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2533.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/459.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/506.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2866.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2645.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/785.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/494.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/500.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3577.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/333.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3205.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3173.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/505.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/397.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/497.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2679.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/481.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/486.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2447.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2641.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2537.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/482.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/487.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/485.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/468.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/498.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2522.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2862.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/484.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/447.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2889.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3375.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2721.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/479.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/495.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2731.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/810.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2725.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1833.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/493.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/568.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/2.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/40.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/747.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1204.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1770.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1002.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3199.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/431.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1360.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1189.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3361.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/964.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1358.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1003.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1599.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/408.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1758.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/225.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/219.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1406.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/351.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1607.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/231.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/187.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/178.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/296.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/190.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/153.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1767.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1573.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/866.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2655.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/444.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3371.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3759.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3749.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/419.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/596.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3413.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1821.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/794.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1006.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/6.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/974.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3349.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/152.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/595.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/204.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/804.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/45.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/392.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/579.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1116.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/46.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1097.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/778.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/557.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/615.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/299.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/423.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/673.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/853.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/224.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1558.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/393.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1035.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/672.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1189.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/736.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/837.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1048.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/569.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1162.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1559.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/554.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/722.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1057.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/623.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/353.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/537.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/709.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/541.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/338.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/779.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/916.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1370.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/390.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1572.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1166.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1604.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1565.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/579.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1360.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/583.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/568.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/637.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/588.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1764.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/582.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1172.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/454.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1600.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/458.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1614.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/401.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/450.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/451.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1003.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1364.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1229.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/413.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/463.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/418.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/456.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/428.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/376.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/444.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/411.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3201.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/461.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/389.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/377.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/400.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/396.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/442.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/427.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2916.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/410.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/398.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/441.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/445.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3215.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2331.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3761.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2484.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1203.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1388.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1407.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1163.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/421.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2494.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2692.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3359.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2335.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/449.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2902.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/141.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3211.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/155.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3239.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3007.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1827.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3017.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1572.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/186.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1389.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3188.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3167.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/50.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3567.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/145.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2480.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3614.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/1374.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/14.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1175.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/15.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3365.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1214.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/5.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/421.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/384.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3003.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/4.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/380.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/24.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2906.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2696.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/347.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/9.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2912.png  \n",
            "  inflating: Face Mask Dataset/Test/WithMask/1174.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/61.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/23.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1759.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/10.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/540.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/18.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/343.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3417.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1210.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/394.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/41.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/376.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/69.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/70.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/383.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/596.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/7.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3172.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/956.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/518.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/597.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/913.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1238.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3013.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1599.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/71.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/561.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/357.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3598.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/924.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/1610.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2453.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/658.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/2669.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/417.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3403.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/321.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/585.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/627.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithoutMask/753.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/329.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/877.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/66.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/280.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/286.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/49.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/464.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3573.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1058.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/293.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/788.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/712.png  \n",
            "  inflating: Face Mask Dataset/Test/WithoutMask/3601.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1054.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/438.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/73.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/562.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/524.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/62.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/8.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/67.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/72.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/36.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/388.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/416.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/17.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/19.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/816.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/55.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/448.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/429.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/37.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/439.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/440.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/414.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/28.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/462.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/374.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/612.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/59.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/505.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/395.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/890.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/381.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/426.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/326.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/32.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/499.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/957.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/412.png  \n",
            "  inflating: Face Mask Dataset/Validation/WithMask/12.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/559.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/646.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/928.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/912.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/697.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/689.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/411.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/433.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/430.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/529.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/300.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1008.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/365.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/664.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/443.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/466.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/934.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/404.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/315.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/383.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/399.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/415.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/446.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/496.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/375.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/452.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/460.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/570.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/418.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/298.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/405.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/455.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/407.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/382.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/340.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/402.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/316.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/417.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/312.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/371.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/700.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/403.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/457.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/358.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/432.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/433.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/465.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/355.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/382.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/406.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/349.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/336.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/300.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/302.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/668.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/325.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/735.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/329.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/304.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/363.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/268.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/846.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/293.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/370.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/341.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/313.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/896.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/310.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/301.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/324.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/932.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/323.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/314.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/764.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/667.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/321.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/895.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/727.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/298.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/335.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/487.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/299.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/367.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1061.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/947.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/403.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/447.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/311.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/317.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/372.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/693.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/707.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/547.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/291.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/412.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1053.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/308.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/845.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/748.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/262.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/356.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/325.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/820.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/342.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/277.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/239.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/339.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/287.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/331.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/927.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/899.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/240.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/599.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/303.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/294.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/256.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/519.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/746.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/250.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/198.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/730.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/270.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/212.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/200.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1079.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/291.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/245.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/889.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/235.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/309.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/188.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/245.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/241.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/194.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/873.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/257.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/229.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/360.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/199.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/462.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/211.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/221.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/202.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/158.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/200.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/271.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/758.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/110.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/244.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1012.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/206.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/327.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/917.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/236.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/398.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/279.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/318.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/330.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/201.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/919.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/287.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/789.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/234.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/337.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/122.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/154.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/272.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/332.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/333.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/368.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/337.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/175.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/314.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/160.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/296.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/120.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/216.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/760.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/125.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/248.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/326.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/388.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/284.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/968.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/399.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/320.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/414.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1076.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/288.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/181.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/195.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/739.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/246.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/323.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/767.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/297.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/217.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/309.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/305.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/809.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/266.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/183.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/159.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/205.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/162.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/361.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/105.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/640.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/259.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/165.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/319.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/290.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/131.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/129.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/108.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/263.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/167.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/289.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/258.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/823.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/229.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/214.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/264.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/629.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/842.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/514.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/113.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/322.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/134.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/156.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/269.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/253.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/334.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/280.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/242.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/101.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/734.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/198.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/139.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/118.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/268.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/104.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/170.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/127.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/119.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/243.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/199.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/136.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/507.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/285.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/449.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/171.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/222.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/247.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/500.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/161.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/295.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/204.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/428.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/278.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/251.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/117.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/841.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/366.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/109.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/103.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/163.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/359.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/249.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/265.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/140.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/645.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/364.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/220.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/348.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/102.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/124.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/172.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1081.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/250.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/274.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/283.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/180.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/687.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/209.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/111.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/925.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/197.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/460.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/116.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/902.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/149.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/126.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/369.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/282.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/238.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/106.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/332.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/367.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/263.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/362.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/273.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/207.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/249.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/316.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/260.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/261.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/135.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/424.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/189.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/174.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/275.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/684.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/338.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/107.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/397.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/164.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/294.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/234.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/157.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/373.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/213.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/177.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/228.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/732.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/215.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/365.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/918.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/286.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/328.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/208.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1127.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/833.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/132.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/278.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/143.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/631.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/236.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/148.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/123.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/288.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1075.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/203.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/166.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/716.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/799.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/354.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/173.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/133.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/176.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/115.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/128.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/649.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/276.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/285.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/713.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/112.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/194.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/777.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1078.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/600.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/100.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/168.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/560.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/130.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/511.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/584.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/253.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/312.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/306.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/494.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/949.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/206.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/138.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1050.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/214.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/886.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/721.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/654.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/523.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/324.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/695.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/578.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/348.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/650.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/808.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/606.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/894.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1007.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/619.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/195.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1025.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/598.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/416.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/441.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/818.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/515.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/671.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/930.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/439.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/142.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/759.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1126.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1085.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/114.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/279.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/527.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/776.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/381.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/708.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1120.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/466.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/212.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/415.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/183.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1074.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1004.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/513.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/289.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/305.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1018.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/215.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/202.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/490.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/616.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/900.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/868.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/939.png  \n",
            "  inflating: Face Mask Dataset/Train/WithoutMask/121.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/872.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/442.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/626.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/869.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/628.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/209.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/717.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/509.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/457.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1030.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1045.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/243.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/549.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/174.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/962.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/331.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/573.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/448.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/738.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/986.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/871.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/642.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/272.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/714.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/1114.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/724.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/465.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/450.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/644.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/641.png  \n",
            "  inflating: Face Mask Dataset/Train/WithMask/647.png  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wHQ3POCK2klw"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "from matplotlib import image\n",
        "\n",
        "from os import listdir\n",
        "from os.path import isfile, join\n",
        "\n",
        "import tensorflow as tf\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras import callbacks, Sequential\n",
        "from tensorflow.keras.layers import Dense, Dropout, Flatten, Conv2D, GlobalAveragePooling2D\n",
        "from keras.applications.vgg16 import VGG16"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y0L42rc7RbKT"
      },
      "source": [
        "# I - <a name=\"preparation\"></a> Traitement du dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_8D71GJklQz9"
      },
      "source": [
        "## A - <a name=\"affichage\"></a> Affichage des images\n",
        "\n",
        "- (a) Exécuter la cellule suivante pour stocker dans des listes ```paths_with_mask``` et ```paths_without_mask``` les chemins des images du dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vk2IuLJ2Rre2"
      },
      "source": [
        "with_mask = 'Face Mask Dataset/Train/WithMask'\n",
        "without_mask = 'Face Mask Dataset/Train/WithoutMask'\n",
        "\n",
        "paths_with_mask = [join(with_mask, f) for f in listdir(with_mask) if isfile(join(with_mask, f))]\n",
        "paths_without_mask = [join(without_mask, f) for f in listdir(without_mask) if isfile(join(without_mask, f))]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DN1hMnP3lxOk"
      },
      "source": [
        "- (b) Utiliser les listes qui viennent d'être définies pour afficher des images d'indice aléatoire.\n",
        "> - La fonction ```image.imread``` permet de transformer un lien en image, qui peut ensuite être affichée par ```plt.imshow```"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "id": "ARRtbN_0Rrw6",
        "outputId": "1e658e98-5af6-4884-90bb-538dbf35fb89"
      },
      "source": [
        "plt.figure(figsize=(6,6))\n",
        "\n",
        "plt.subplot(221)\n",
        "plt.axis('off')\n",
        "plt.imshow(image.imread(paths_with_mask[np.random.randint(len(paths_with_mask))]))\n",
        "\n",
        "plt.subplot(222)\n",
        "plt.axis('off')\n",
        "plt.imshow(image.imread(paths_without_mask[np.random.randint(len(paths_without_mask))]))\n",
        "\n",
        "plt.subplot(223)\n",
        "plt.axis('off')\n",
        "plt.imshow(image.imread(paths_with_mask[np.random.randint(len(paths_with_mask))]))\n",
        "\n",
        "plt.subplot(224)\n",
        "plt.axis('off')\n",
        "plt.imshow(image.imread(paths_without_mask[np.random.randint(len(paths_without_mask))]));"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uhTP03RxRbQp"
      },
      "source": [
        "## B - <a name=\"generateurs\"></a> Générateurs\n",
        "\n",
        "- (c) Définir trois générateurs :\n",
        "> - ```train_data_generator``` avec les paramètres de transformation de votre choix (on prendra un ```validation_split``` de 0.2 et un scaling de 1/255)\n",
        ">\n",
        "> - ```valid_data_generator``` avec un ```validation_split``` de 0.2 et un scaling de 1/255\n",
        ">\n",
        "> - ```test_data_generator``` avec un scaling de 1/255"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "08km-K7Q2zuF"
      },
      "source": [
        "train_data_generator = ImageDataGenerator(rescale = 1./255,\n",
        "                                   validation_split = 0.2,\n",
        "                                   rotation_range=5,\n",
        "                                   width_shift_range=0.2,\n",
        "                                   height_shift_range=0.2,\n",
        "                                   shear_range=0.2,\n",
        "                                   zoom_range=0.2,\n",
        "                                   horizontal_flip=True)\n",
        "\n",
        "valid_data_generator = ImageDataGenerator(rescale = 1./255,\n",
        "                                   validation_split = 0.2)\n",
        "\n",
        "test_data_generator  = ImageDataGenerator(rescale = 1./255)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "l-drAS4ZnatL"
      },
      "source": [
        "- (d) Implémenter sous les noms ```training_data```, ```valid_data``` et ```test_data``` des itérateurs des objets ```train_data_generator```, ```valid_data_generator``` et ```test_data_generator``` (utiliser la méthode ```flow```).\n",
        "> - Les répertoires à utiliser sont ```'Face Mask Dataset/Train'```, ```'Face Mask Dataset/Validation'``` et ```'Face Mask Dataset/Test'```\n",
        ">\n",
        "> - On choisira une taille cible de (224, 224) (pour le ```VGG16```), une classe binaire (avec ou sans masque), et des lots de données de taille 32 (pour train et valid) ou 1 (pour test)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kRbXLoMv29mR",
        "outputId": "276fa6dc-b5d6-4dee-c788-1e0a09d94522"
      },
      "source": [
        "training_data  = train_data_generator.flow_from_directory(directory = 'Face Mask Dataset/Train',\n",
        "                                                   target_size = (224, 224),\n",
        "                                                   class_mode = 'binary',\n",
        "                                                   batch_size = 32)\n",
        "\n",
        "valid_data = valid_data_generator.flow_from_directory(directory = 'Face Mask Dataset/Validation',\n",
        "                                                  target_size = (224, 224),\n",
        "                                                  class_mode = 'binary',\n",
        "                                                  batch_size = 32)\n",
        "\n",
        "test_data = test_data_generator.flow_from_directory(directory = 'Face Mask Dataset/Test',\n",
        "                                                  target_size = (224, 224),\n",
        "                                                  class_mode = 'binary',\n",
        "                                                  batch_size = 1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 654 images belonging to 2 classes.\n",
            "Found 183 images belonging to 2 classes.\n",
            "Found 166 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OOHwqp7eomul"
      },
      "source": [
        "# II - <a name=\"classif\"></a> Classification\n",
        "\n",
        "- (a) Importer le modèle ```VGG16``` et freezer ses couches.\n",
        "\n",
        "- (b) Instancier un réseau de neurones (appelé ```model```), lui ajouter le modèle ```VGG16``` et ajouter les couches denses de votre choix.\n",
        "\n",
        "- (c) Compiler le modèle avec une métrique appropriée.\n",
        "\n",
        "- (d) Entraîner le modèle avec les paramètres de votre choix. On pourra également définir des callbacks.\n",
        "\n",
        "- (e) Evaluer le modèle sur les données de test avec la méthode ```evaluate```."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CIFvE3eA3KqA",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "76da7442-21e8-40f6-ba43-6e39ac486c24"
      },
      "source": [
        "# Modèle VGG16\n",
        "base_model = VGG16(weights='imagenet', include_top=False)\n",
        "# Freezer les couches du VGG16\n",
        "for layer in base_model.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "model = Sequential()\n",
        "model.add(base_model) # Ajout du modèle VGG16\n",
        "model.add(GlobalAveragePooling2D())\n",
        "model.add(Flatten())\n",
        "model.add(Dropout(rate=0.2))\n",
        "model.add(Dense(units=32, activation='relu'))\n",
        "model.add(Dense(units=32, activation='relu'))\n",
        "model.add(Dropout(rate=0.2))\n",
        "model.add(Dense(units=32, activation='relu'))\n",
        "model.add(Dense(1, activation='sigmoid'))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "58892288/58889256 [==============================] - 2s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uXwz_X0t8Bj6"
      },
      "source": [
        "early_stopping = callbacks.EarlyStopping(monitor = 'val_loss',\n",
        "                                         patience = 4,\n",
        "                                         mode = 'min',\n",
        "                                         restore_best_weights = True)\n",
        "\n",
        "lr_plateau = callbacks.ReduceLROnPlateau(monitor = 'val_loss',\n",
        "                                        patience=3,\n",
        "                                         factor=0.5,\n",
        "                                         verbose=2,\n",
        "                                         mode='min',\n",
        "                                         min_lr = 1e-10)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P4mvz9228DOi"
      },
      "source": [
        "model.compile(optimizer='Adam', loss='binary_crossentropy',metrics=['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ze0BJmbB8E97",
        "outputId": "d7087b89-89bf-42c4-e6c0-6b2e8946f136"
      },
      "source": [
        "history = model.fit(training_data, validation_data=valid_data, epochs=10, callbacks=[early_stopping,lr_plateau])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "21/21 [==============================] - 417s 20s/step - loss: 0.6884 - accuracy: 0.5306 - val_loss: 0.6300 - val_accuracy: 0.9126\n",
            "Epoch 2/10\n",
            "21/21 [==============================] - 413s 20s/step - loss: 0.6366 - accuracy: 0.7396 - val_loss: 0.5258 - val_accuracy: 0.8197\n",
            "Epoch 3/10\n",
            "21/21 [==============================] - 413s 20s/step - loss: 0.5279 - accuracy: 0.8130 - val_loss: 0.3266 - val_accuracy: 0.9016\n",
            "Epoch 4/10\n",
            "21/21 [==============================] - 413s 20s/step - loss: 0.3517 - accuracy: 0.8902 - val_loss: 0.1955 - val_accuracy: 0.9399\n",
            "Epoch 5/10\n",
            "21/21 [==============================] - 412s 20s/step - loss: 0.2342 - accuracy: 0.9092 - val_loss: 0.1512 - val_accuracy: 0.9508\n",
            "Epoch 6/10\n",
            "21/21 [==============================] - 414s 20s/step - loss: 0.1852 - accuracy: 0.9307 - val_loss: 0.1288 - val_accuracy: 0.9563\n",
            "Epoch 7/10\n",
            "21/21 [==============================] - 415s 20s/step - loss: 0.2011 - accuracy: 0.9210 - val_loss: 0.1207 - val_accuracy: 0.9563\n",
            "Epoch 8/10\n",
            "21/21 [==============================] - 414s 20s/step - loss: 0.2006 - accuracy: 0.9273 - val_loss: 0.1104 - val_accuracy: 0.9563\n",
            "Epoch 9/10\n",
            "21/21 [==============================] - 412s 20s/step - loss: 0.1751 - accuracy: 0.9379 - val_loss: 0.1201 - val_accuracy: 0.9563\n",
            "Epoch 10/10\n",
            "21/21 [==============================] - 413s 20s/step - loss: 0.1516 - accuracy: 0.9518 - val_loss: 0.1063 - val_accuracy: 0.9617\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "9zMJnvDfTEVU",
        "outputId": "03c4c0be-773f-4f46-9356-a0caa2290a59"
      },
      "source": [
        "plt.figure(figsize=(12,4))\n",
        "plt.subplot(121)\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('Model loss by epoch')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='right')\n",
        "\n",
        "plt.subplot(122)\n",
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.title('Model acc by epoch')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='right')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZipeKpavi0z-",
        "outputId": "37d563ae-688d-4f80-825b-7867e63cb307"
      },
      "source": [
        "model.evaluate(test_data)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "166/166 [==============================] - 86s 518ms/step - loss: 0.1543 - accuracy: 0.9217\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.15433886647224426, 0.9216867685317993]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OJ0rkEHOkD2m"
      },
      "source": [
        "# III - <a name=\"test\"></a> Test en temps réel\n",
        "\n",
        "- (a) Lancer les cellules suivantes pour tester votre modèle en direct sur votre webcam."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r4K5NQogluPU"
      },
      "source": [
        "labels_dict = {1 : 'No mask', 0 : 'Mask'}"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I43WOiYHlw42"
      },
      "source": [
        "from IPython.display import display, Javascript, Image\n",
        "from google.colab.output import eval_js\n",
        "from base64 import b64decode, b64encode\n",
        "import cv2\n",
        "import numpy as np\n",
        "import PIL\n",
        "import io\n",
        "import html\n",
        "import time\n",
        "\n",
        "def js_to_image(js_reply):\n",
        "  image_bytes = b64decode(js_reply.split(',')[1])\n",
        "  jpg_as_np = np.frombuffer(image_bytes, dtype=np.uint8)\n",
        "  img = cv2.imdecode(jpg_as_np, flags=1)\n",
        "  return img\n",
        "\n",
        "def bbox_to_bytes(bbox_array):\n",
        "  bbox_PIL = PIL.Image.fromarray(bbox_array, 'RGBA')\n",
        "  iobuf = io.BytesIO()\n",
        "  bbox_PIL.save(iobuf, format='png')\n",
        "  bbox_bytes = 'data:image/png;base64,{}'.format((str(b64encode(iobuf.getvalue()), 'utf-8')))\n",
        "  return bbox_bytes\n",
        "\n",
        "def video_stream():\n",
        "  js = Javascript('''\n",
        "    var video;\n",
        "    var div = null;\n",
        "    var stream;\n",
        "    var captureCanvas;\n",
        "    var imgElement;\n",
        "    var labelElement;\n",
        "\n",
        "    var pendingResolve = null;\n",
        "    var shutdown = false;\n",
        "\n",
        "    function removeDom() {\n",
        "       stream.getVideoTracks()[0].stop();\n",
        "       video.remove();\n",
        "       div.remove();\n",
        "       video = null;\n",
        "       div = null;\n",
        "       stream = null;\n",
        "       imgElement = null;\n",
        "       captureCanvas = null;\n",
        "       labelElement = null;\n",
        "    }\n",
        "\n",
        "    function onAnimationFrame() {\n",
        "      if (!shutdown) {\n",
        "        window.requestAnimationFrame(onAnimationFrame);\n",
        "      }\n",
        "      if (pendingResolve) {\n",
        "        var result = \"\";\n",
        "        if (!shutdown) {\n",
        "          captureCanvas.getContext('2d').drawImage(video, 0, 0, 640, 480);\n",
        "          result = captureCanvas.toDataURL('image/jpeg', 0.8)\n",
        "        }\n",
        "        var lp = pendingResolve;\n",
        "        pendingResolve = null;\n",
        "        lp(result);\n",
        "      }\n",
        "    }\n",
        "\n",
        "    async function createDom() {\n",
        "      if (div !== null) {\n",
        "        return stream;\n",
        "      }\n",
        "\n",
        "      div = document.createElement('div');\n",
        "      div.style.border = '2px solid black';\n",
        "      div.style.padding = '3px';\n",
        "      div.style.width = '100%';\n",
        "      div.style.maxWidth = '600px';\n",
        "      document.body.appendChild(div);\n",
        "\n",
        "      const modelOut = document.createElement('div');\n",
        "      modelOut.innerHTML = \"<span>Status:</span>\";\n",
        "      labelElement = document.createElement('span');\n",
        "      labelElement.innerText = 'No data';\n",
        "      labelElement.style.fontWeight = 'bold';\n",
        "      modelOut.appendChild(labelElement);\n",
        "      div.appendChild(modelOut);\n",
        "\n",
        "      video = document.createElement('video');\n",
        "      video.style.display = 'block';\n",
        "      video.width = div.clientWidth - 6;\n",
        "      video.setAttribute('playsinline', '');\n",
        "      video.onclick = () => { shutdown = true; };\n",
        "      stream = await navigator.mediaDevices.getUserMedia(\n",
        "          {video: { facingMode: \"environment\"}});\n",
        "      div.appendChild(video);\n",
        "\n",
        "      imgElement = document.createElement('img');\n",
        "      imgElement.style.position = 'absolute';\n",
        "      imgElement.style.zIndex = 1;\n",
        "      imgElement.onclick = () => { shutdown = true; };\n",
        "      div.appendChild(imgElement);\n",
        "\n",
        "      const instruction = document.createElement('div');\n",
        "      instruction.innerHTML =\n",
        "          '<span style=\"color: red; font-weight: bold;\">' +\n",
        "          'When finished, click here or on the video to stop this demo</span>';\n",
        "      div.appendChild(instruction);\n",
        "      instruction.onclick = () => { shutdown = true; };\n",
        "\n",
        "      video.srcObject = stream;\n",
        "      await video.play();\n",
        "\n",
        "      captureCanvas = document.createElement('canvas');\n",
        "      captureCanvas.width = 640; //video.videoWidth;\n",
        "      captureCanvas.height = 480; //video.videoHeight;\n",
        "      window.requestAnimationFrame(onAnimationFrame);\n",
        "\n",
        "      return stream;\n",
        "    }\n",
        "    async function stream_frame(label, imgData) {\n",
        "      if (shutdown) {\n",
        "        removeDom();\n",
        "        shutdown = false;\n",
        "        return '';\n",
        "      }\n",
        "\n",
        "      var preCreate = Date.now();\n",
        "      stream = await createDom();\n",
        "\n",
        "      var preShow = Date.now();\n",
        "      if (label != \"\") {\n",
        "        labelElement.innerHTML = label;\n",
        "      }\n",
        "\n",
        "      if (imgData != \"\") {\n",
        "        var videoRect = video.getClientRects()[0];\n",
        "        imgElement.style.top = videoRect.top + \"px\";\n",
        "        imgElement.style.left = videoRect.left + \"px\";\n",
        "        imgElement.style.width = videoRect.width + \"px\";\n",
        "        imgElement.style.height = videoRect.height + \"px\";\n",
        "        imgElement.src = imgData;\n",
        "      }\n",
        "\n",
        "      var preCapture = Date.now();\n",
        "      var result = await new Promise(function(resolve, reject) {\n",
        "        pendingResolve = resolve;\n",
        "      });\n",
        "      shutdown = false;\n",
        "\n",
        "      return {'create': preShow - preCreate,\n",
        "              'show': preCapture - preShow,\n",
        "              'capture': Date.now() - preCapture,\n",
        "              'img': result};\n",
        "    }\n",
        "    ''')\n",
        "\n",
        "  display(js)\n",
        "\n",
        "def video_frame(label, bbox):\n",
        "  data = eval_js('stream_frame(\"{}\", \"{}\")'.format(label, bbox))\n",
        "  return data"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "id": "HnSyL6yjkOZI",
        "outputId": "7ef5b30b-b2b9-4b8d-dee8-c1786d342e41"
      },
      "source": [
        "video_stream()\n",
        "label_html = 'Capturing...'\n",
        "bbox = ''\n",
        "count = 0\n",
        "\n",
        "while True:\n",
        "    js_reply = video_frame(label_html, bbox)\n",
        "    if not js_reply:\n",
        "        break\n",
        "    img = js_to_image(js_reply[\"img\"])\n",
        "    bbox_array = np.zeros([480,640,4], dtype=np.uint8)\n",
        "    x, y, w, h = 220, 165, 200, 220\n",
        "    face_img=img[y:y+w,x:x+w]\n",
        "    resized=cv2.resize(face_img,(224,224))\n",
        "    reshaped=np.reshape(resized,(1,224,224,3))\n",
        "    result=model.predict(reshaped).round()[0][0]\n",
        "    if result == 0:\n",
        "      bbox_array = cv2.rectangle(bbox_array,(x,y),(x+w,y+h),(0,255,0),2)\n",
        "      cv2.putText(bbox_array, labels_dict[result], (x, y-10),cv2.FONT_HERSHEY_SIMPLEX,0.8,(0,255,0),2)\n",
        "    if result == 1:\n",
        "      bbox_array = cv2.rectangle(bbox_array,(x,y),(x+w,y+h),(255,0,0),2)\n",
        "      cv2.putText(bbox_array, labels_dict[result], (x, y-10),cv2.FONT_HERSHEY_SIMPLEX,0.8,(255,0,0),2)\n",
        "\n",
        "    bbox_array[:,:,3] = (bbox_array.max(axis = 2) > 0 ).astype(int) * 255\n",
        "    bbox_bytes = bbox_to_bytes(bbox_array)\n",
        "    bbox = bbox_bytes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "\n",
              "    var video;\n",
              "    var div = null;\n",
              "    var stream;\n",
              "    var captureCanvas;\n",
              "    var imgElement;\n",
              "    var labelElement;\n",
              "    \n",
              "    var pendingResolve = null;\n",
              "    var shutdown = false;\n",
              "    \n",
              "    function removeDom() {\n",
              "       stream.getVideoTracks()[0].stop();\n",
              "       video.remove();\n",
              "       div.remove();\n",
              "       video = null;\n",
              "       div = null;\n",
              "       stream = null;\n",
              "       imgElement = null;\n",
              "       captureCanvas = null;\n",
              "       labelElement = null;\n",
              "    }\n",
              "    \n",
              "    function onAnimationFrame() {\n",
              "      if (!shutdown) {\n",
              "        window.requestAnimationFrame(onAnimationFrame);\n",
              "      }\n",
              "      if (pendingResolve) {\n",
              "        var result = \"\";\n",
              "        if (!shutdown) {\n",
              "          captureCanvas.getContext('2d').drawImage(video, 0, 0, 640, 480);\n",
              "          result = captureCanvas.toDataURL('image/jpeg', 0.8)\n",
              "        }\n",
              "        var lp = pendingResolve;\n",
              "        pendingResolve = null;\n",
              "        lp(result);\n",
              "      }\n",
              "    }\n",
              "    \n",
              "    async function createDom() {\n",
              "      if (div !== null) {\n",
              "        return stream;\n",
              "      }\n",
              "\n",
              "      div = document.createElement('div');\n",
              "      div.style.border = '2px solid black';\n",
              "      div.style.padding = '3px';\n",
              "      div.style.width = '100%';\n",
              "      div.style.maxWidth = '600px';\n",
              "      document.body.appendChild(div);\n",
              "      \n",
              "      const modelOut = document.createElement('div');\n",
              "      modelOut.innerHTML = \"<span>Status:</span>\";\n",
              "      labelElement = document.createElement('span');\n",
              "      labelElement.innerText = 'No data';\n",
              "      labelElement.style.fontWeight = 'bold';\n",
              "      modelOut.appendChild(labelElement);\n",
              "      div.appendChild(modelOut);\n",
              "           \n",
              "      video = document.createElement('video');\n",
              "      video.style.display = 'block';\n",
              "      video.width = div.clientWidth - 6;\n",
              "      video.setAttribute('playsinline', '');\n",
              "      video.onclick = () => { shutdown = true; };\n",
              "      stream = await navigator.mediaDevices.getUserMedia(\n",
              "          {video: { facingMode: \"environment\"}});\n",
              "      div.appendChild(video);\n",
              "\n",
              "      imgElement = document.createElement('img');\n",
              "      imgElement.style.position = 'absolute';\n",
              "      imgElement.style.zIndex = 1;\n",
              "      imgElement.onclick = () => { shutdown = true; };\n",
              "      div.appendChild(imgElement);\n",
              "      \n",
              "      const instruction = document.createElement('div');\n",
              "      instruction.innerHTML = \n",
              "          '<span style=\"color: red; font-weight: bold;\">' +\n",
              "          'When finished, click here or on the video to stop this demo</span>';\n",
              "      div.appendChild(instruction);\n",
              "      instruction.onclick = () => { shutdown = true; };\n",
              "      \n",
              "      video.srcObject = stream;\n",
              "      await video.play();\n",
              "\n",
              "      captureCanvas = document.createElement('canvas');\n",
              "      captureCanvas.width = 640; //video.videoWidth;\n",
              "      captureCanvas.height = 480; //video.videoHeight;\n",
              "      window.requestAnimationFrame(onAnimationFrame);\n",
              "      \n",
              "      return stream;\n",
              "    }\n",
              "    async function stream_frame(label, imgData) {\n",
              "      if (shutdown) {\n",
              "        removeDom();\n",
              "        shutdown = false;\n",
              "        return '';\n",
              "      }\n",
              "\n",
              "      var preCreate = Date.now();\n",
              "      stream = await createDom();\n",
              "      \n",
              "      var preShow = Date.now();\n",
              "      if (label != \"\") {\n",
              "        labelElement.innerHTML = label;\n",
              "      }\n",
              "            \n",
              "      if (imgData != \"\") {\n",
              "        var videoRect = video.getClientRects()[0];\n",
              "        imgElement.style.top = videoRect.top + \"px\";\n",
              "        imgElement.style.left = videoRect.left + \"px\";\n",
              "        imgElement.style.width = videoRect.width + \"px\";\n",
              "        imgElement.style.height = videoRect.height + \"px\";\n",
              "        imgElement.src = imgData;\n",
              "      }\n",
              "      \n",
              "      var preCapture = Date.now();\n",
              "      var result = await new Promise(function(resolve, reject) {\n",
              "        pendingResolve = resolve;\n",
              "      });\n",
              "      shutdown = false;\n",
              "      \n",
              "      return {'create': preShow - preCreate, \n",
              "              'show': preCapture - preShow, \n",
              "              'capture': Date.now() - preCapture,\n",
              "              'img': result};\n",
              "    }\n",
              "    "
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}