FaceR&V.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "FaceR&V",
      "provenance": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyOkJmi50X7SdLnnUNBKlvdI"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "CTNQAdD8Dd1X",
        "outputId": "bbfeb036-917e-4103-ebf0-36def65f6a0e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "import sys\n",
        "sys.path"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['',\n",
              " '/env/python',\n",
              " '/usr/lib/python36.zip',\n",
              " '/usr/lib/python3.6',\n",
              " '/usr/lib/python3.6/lib-dynload',\n",
              " '/usr/local/lib/python3.6/dist-packages',\n",
              " '/usr/lib/python3/dist-packages',\n",
              " '/usr/local/lib/python3.6/dist-packages/IPython/extensions',\n",
              " '/root/.ipython']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 1
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uNM2wj4SriVK"
      },
      "source": [
        "# **Uploading files from machine**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nRIheMQlIEnP",
        "outputId": "f166a9de-60d6-473e-cddd-5f3032365d4d",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 107
        }
      },
      "source": [
        "from google.colab import files\n",
        "uploaded = files.upload()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-d99d3620-5088-47ad-9e70-841f3608cec6\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-d99d3620-5088-47ad-9e70-841f3608cec6\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving meOne.jpg to meOne.jpg\n",
            "Saving maOne.jpg to maOne.jpg\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MXdXOTdJsEoy"
      },
      "source": [
        "Using tensorfow version 1 in order to match compatibility"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BGSQI0GNYqhc",
        "outputId": "e2b9a6e4-81d2-46c0-961b-709b4c48b4ab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "import tensorflow.compat.v1 as tf\n",
        "tf.disable_v2_behavior()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/compat/v2_compat.py:96: disable_resource_variables (from tensorflow.python.ops.variable_scope) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "non-resource variables are not supported in the long term\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tt67rn-QYMXg"
      },
      "source": [
        "#import tensorflow as tf\n",
        "\n",
        "tf.enable_eager_execution()\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J78z_msZ4lA2"
      },
      "source": [
        "."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MbUsVs5lYg4-",
        "outputId": "7dbc4887-3dab-4d5f-dfcd-9cdefca92c63",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tf.executing_eagerly()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vEJGNyZhsP1Q"
      },
      "source": [
        "# Importing required libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OxneTru-It-g"
      },
      "source": [
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D, ZeroPadding2D, Activation, Input, concatenate\n",
        "from keras.models import Model\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "from keras.layers.pooling import MaxPooling2D, AveragePooling2D\n",
        "from keras.layers.merge import Concatenate\n",
        "from keras.layers.core import Lambda, Flatten, Dense\n",
        "from keras.initializers import glorot_uniform\n",
        "from keras.engine.topology import Layer\n",
        "from keras import backend as K\n",
        "K.set_image_data_format('channels_first')\n",
        "import cv2\n",
        "import os\n",
        "import numpy as np\n",
        "from numpy import genfromtxt\n",
        "import pandas as pd\n",
        "\n",
        "%matplotlib inline\n",
        "%load_ext autoreload\n",
        "%autoreload 2\n",
        "\n",
        "#np.set_printoptions(threshold=np.nan)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y-kyq4zHtGSH"
      },
      "source": [
        "......"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3OYAy40UtGAF"
      },
      "source": [
        "......"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UnaCFlHrtFsQ"
      },
      "source": [
        "......"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d6MuJpYcsaDZ"
      },
      "source": [
        "Importing fr_utils and the inception_block_v2 architecture which has already been imported from the machine to Colab"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gjy9cv3vYQlf"
      },
      "source": [
        "\n",
        "from fr_utils import *\n",
        "from inception_blocks_v2 import *\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BYfXUfjQtMMV"
      },
      "source": [
        "......"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zlaTBgYEtL2q"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "POQeazPMtLGk"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "viYLq9rgT50O"
      },
      "source": [
        "# 1st Model (Inception V2)\n",
        "\n",
        "Model summary"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ck4hfNB7JBFR",
        "outputId": "2be0a2ec-9927-4f28-f00b-87fabc9846c8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "FRmodel = faceRecoModel(input_shape=(3, 96, 96))\n",
        "FRmodel.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"FaceRecoModel\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 3, 96, 96)]  0                                            \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d (ZeroPadding2D)  (None, 3, 102, 102)  0           input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1 (Conv2D)                  (None, 64, 48, 48)   9472        zero_padding2d[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "bn1 (BatchNormalization)        (None, 64, 48, 48)   256         conv1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation (Activation)         (None, 64, 48, 48)   0           bn1[0][0]                        \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_1 (ZeroPadding2D (None, 64, 50, 50)   0           activation[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D)    (None, 64, 24, 24)   0           zero_padding2d_1[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2 (Conv2D)                  (None, 64, 24, 24)   4160        max_pooling2d[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "bn2 (BatchNormalization)        (None, 64, 24, 24)   256         conv2[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_1 (Activation)       (None, 64, 24, 24)   0           bn2[0][0]                        \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_2 (ZeroPadding2D (None, 64, 26, 26)   0           activation_1[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv3 (Conv2D)                  (None, 192, 24, 24)  110784      zero_padding2d_2[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "bn3 (BatchNormalization)        (None, 192, 24, 24)  768         conv3[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_2 (Activation)       (None, 192, 24, 24)  0           bn3[0][0]                        \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_3 (ZeroPadding2D (None, 192, 26, 26)  0           activation_2[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2D)  (None, 192, 12, 12)  0           zero_padding2d_3[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_3x3_conv1 (Conv2D) (None, 96, 12, 12)   18528       max_pooling2d_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_5x5_conv1 (Conv2D) (None, 16, 12, 12)   3088        max_pooling2d_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_3x3_bn1 (BatchNorm (None, 96, 12, 12)   384         inception_3a_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_5x5_bn1 (BatchNorm (None, 16, 12, 12)   64          inception_3a_5x5_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_3 (Activation)       (None, 96, 12, 12)   0           inception_3a_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_5 (Activation)       (None, 16, 12, 12)   0           inception_3a_5x5_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2D)  (None, 192, 5, 5)    0           max_pooling2d_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_4 (ZeroPadding2D (None, 96, 14, 14)   0           activation_3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_5 (ZeroPadding2D (None, 16, 16, 16)   0           activation_5[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_pool_conv (Conv2D) (None, 32, 5, 5)     6176        max_pooling2d_2[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_3x3_conv2 (Conv2D) (None, 128, 12, 12)  110720      zero_padding2d_4[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_5x5_conv2 (Conv2D) (None, 32, 12, 12)   12832       zero_padding2d_5[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_pool_bn (BatchNorm (None, 32, 5, 5)     128         inception_3a_pool_conv[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_1x1_conv (Conv2D)  (None, 64, 12, 12)   12352       max_pooling2d_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_3x3_bn2 (BatchNorm (None, 128, 12, 12)  512         inception_3a_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_5x5_bn2 (BatchNorm (None, 32, 12, 12)   128         inception_3a_5x5_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_7 (Activation)       (None, 32, 5, 5)     0           inception_3a_pool_bn[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "inception_3a_1x1_bn (BatchNorma (None, 64, 12, 12)   256         inception_3a_1x1_conv[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_4 (Activation)       (None, 128, 12, 12)  0           inception_3a_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_6 (Activation)       (None, 32, 12, 12)   0           inception_3a_5x5_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_6 (ZeroPadding2D (None, 32, 12, 12)   0           activation_7[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_8 (Activation)       (None, 64, 12, 12)   0           inception_3a_1x1_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "concatenate (Concatenate)       (None, 256, 12, 12)  0           activation_4[0][0]               \n",
            "                                                                 activation_6[0][0]               \n",
            "                                                                 zero_padding2d_6[0][0]           \n",
            "                                                                 activation_8[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_3x3_conv1 (Conv2D) (None, 96, 12, 12)   24672       concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_5x5_conv1 (Conv2D) (None, 32, 12, 12)   8224        concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_3x3_bn1 (BatchNorm (None, 96, 12, 12)   384         inception_3b_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_5x5_bn1 (BatchNorm (None, 32, 12, 12)   128         inception_3b_5x5_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_9 (Activation)       (None, 96, 12, 12)   0           inception_3b_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_11 (Activation)      (None, 32, 12, 12)   0           inception_3b_5x5_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d (AveragePooli (None, 256, 4, 4)    0           concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_7 (ZeroPadding2D (None, 96, 14, 14)   0           activation_9[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_8 (ZeroPadding2D (None, 32, 16, 16)   0           activation_11[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_pool_conv (Conv2D) (None, 64, 4, 4)     16448       average_pooling2d[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_3x3_conv2 (Conv2D) (None, 128, 12, 12)  110720      zero_padding2d_7[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_5x5_conv2 (Conv2D) (None, 64, 12, 12)   51264       zero_padding2d_8[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_pool_bn (BatchNorm (None, 64, 4, 4)     256         inception_3b_pool_conv[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_1x1_conv (Conv2D)  (None, 64, 12, 12)   16448       concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_3x3_bn2 (BatchNorm (None, 128, 12, 12)  512         inception_3b_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_5x5_bn2 (BatchNorm (None, 64, 12, 12)   256         inception_3b_5x5_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_13 (Activation)      (None, 64, 4, 4)     0           inception_3b_pool_bn[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "inception_3b_1x1_bn (BatchNorma (None, 64, 12, 12)   256         inception_3b_1x1_conv[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_10 (Activation)      (None, 128, 12, 12)  0           inception_3b_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_12 (Activation)      (None, 64, 12, 12)   0           inception_3b_5x5_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_9 (ZeroPadding2D (None, 64, 12, 12)   0           activation_13[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_14 (Activation)      (None, 64, 12, 12)   0           inception_3b_1x1_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 320, 12, 12)  0           activation_10[0][0]              \n",
            "                                                                 activation_12[0][0]              \n",
            "                                                                 zero_padding2d_9[0][0]           \n",
            "                                                                 activation_14[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_3x3_conv1 (Conv2D) (None, 128, 12, 12)  41088       concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_5x5_conv1 (Conv2D) (None, 32, 12, 12)   10272       concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_3x3_bn1 (BatchNorm (None, 128, 12, 12)  512         inception_3c_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_5x5_bn1 (BatchNorm (None, 32, 12, 12)   128         inception_3c_5x5_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_15 (Activation)      (None, 128, 12, 12)  0           inception_3c_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_17 (Activation)      (None, 32, 12, 12)   0           inception_3c_5x5_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_10 (ZeroPadding2 (None, 128, 14, 14)  0           activation_15[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_11 (ZeroPadding2 (None, 32, 16, 16)   0           activation_17[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_3x3_conv2 (Conv2D) (None, 256, 6, 6)    295168      zero_padding2d_10[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_5x5_conv2 (Conv2D) (None, 64, 6, 6)     51264       zero_padding2d_11[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_3x3_bn2 (BatchNorm (None, 256, 6, 6)    1024        inception_3c_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_3c_5x5_bn2 (BatchNorm (None, 64, 6, 6)     256         inception_3c_5x5_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2D)  (None, 320, 5, 5)    0           concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_16 (Activation)      (None, 256, 6, 6)    0           inception_3c_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_18 (Activation)      (None, 64, 6, 6)     0           inception_3c_5x5_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_12 (ZeroPadding2 (None, 320, 6, 6)    0           max_pooling2d_3[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_2 (Concatenate)     (None, 640, 6, 6)    0           activation_16[0][0]              \n",
            "                                                                 activation_18[0][0]              \n",
            "                                                                 zero_padding2d_12[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_3x3_conv1 (Conv2D) (None, 96, 6, 6)     61536       concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_5x5_conv1 (Conv2D) (None, 32, 6, 6)     20512       concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_3x3_bn1 (BatchNorm (None, 96, 6, 6)     384         inception_4a_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_5x5_bn1 (BatchNorm (None, 32, 6, 6)     128         inception_4a_5x5_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_19 (Activation)      (None, 96, 6, 6)     0           inception_4a_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_21 (Activation)      (None, 32, 6, 6)     0           inception_4a_5x5_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_1 (AveragePoo (None, 640, 2, 2)    0           concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_13 (ZeroPadding2 (None, 96, 8, 8)     0           activation_19[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_14 (ZeroPadding2 (None, 32, 10, 10)   0           activation_21[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_pool_conv (Conv2D) (None, 128, 2, 2)    82048       average_pooling2d_1[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_3x3_conv2 (Conv2D) (None, 192, 6, 6)    166080      zero_padding2d_13[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_5x5_conv2 (Conv2D) (None, 64, 6, 6)     51264       zero_padding2d_14[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_pool_bn (BatchNorm (None, 128, 2, 2)    512         inception_4a_pool_conv[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_1x1_conv (Conv2D)  (None, 256, 6, 6)    164096      concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_3x3_bn2 (BatchNorm (None, 192, 6, 6)    768         inception_4a_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_5x5_bn2 (BatchNorm (None, 64, 6, 6)     256         inception_4a_5x5_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_23 (Activation)      (None, 128, 2, 2)    0           inception_4a_pool_bn[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "inception_4a_1x1_bn (BatchNorma (None, 256, 6, 6)    1024        inception_4a_1x1_conv[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_20 (Activation)      (None, 192, 6, 6)    0           inception_4a_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_22 (Activation)      (None, 64, 6, 6)     0           inception_4a_5x5_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_15 (ZeroPadding2 (None, 128, 6, 6)    0           activation_23[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_24 (Activation)      (None, 256, 6, 6)    0           inception_4a_1x1_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_3 (Concatenate)     (None, 640, 6, 6)    0           activation_20[0][0]              \n",
            "                                                                 activation_22[0][0]              \n",
            "                                                                 zero_padding2d_15[0][0]          \n",
            "                                                                 activation_24[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_3x3_conv1 (Conv2D) (None, 160, 6, 6)    102560      concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_5x5_conv1 (Conv2D) (None, 64, 6, 6)     41024       concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_3x3_bn1 (BatchNorm (None, 160, 6, 6)    640         inception_4e_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_5x5_bn1 (BatchNorm (None, 64, 6, 6)     256         inception_4e_5x5_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_25 (Activation)      (None, 160, 6, 6)    0           inception_4e_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_27 (Activation)      (None, 64, 6, 6)     0           inception_4e_5x5_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_16 (ZeroPadding2 (None, 160, 8, 8)    0           activation_25[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_17 (ZeroPadding2 (None, 64, 10, 10)   0           activation_27[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_3x3_conv2 (Conv2D) (None, 256, 3, 3)    368896      zero_padding2d_16[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_5x5_conv2 (Conv2D) (None, 128, 3, 3)    204928      zero_padding2d_17[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_3x3_bn2 (BatchNorm (None, 256, 3, 3)    1024        inception_4e_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_4e_5x5_bn2 (BatchNorm (None, 128, 3, 3)    512         inception_4e_5x5_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2D)  (None, 640, 2, 2)    0           concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_26 (Activation)      (None, 256, 3, 3)    0           inception_4e_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "activation_28 (Activation)      (None, 128, 3, 3)    0           inception_4e_5x5_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_18 (ZeroPadding2 (None, 640, 3, 3)    0           max_pooling2d_4[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_4 (Concatenate)     (None, 1024, 3, 3)   0           activation_26[0][0]              \n",
            "                                                                 activation_28[0][0]              \n",
            "                                                                 zero_padding2d_18[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_3x3_conv1 (Conv2D) (None, 96, 3, 3)     98400       concatenate_4[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_3x3_bn1 (BatchNorm (None, 96, 3, 3)     384         inception_5a_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_29 (Activation)      (None, 96, 3, 3)     0           inception_5a_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_2 (AveragePoo (None, 1024, 1, 1)   0           concatenate_4[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_19 (ZeroPadding2 (None, 96, 5, 5)     0           activation_29[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_pool_conv (Conv2D) (None, 96, 1, 1)     98400       average_pooling2d_2[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_3x3_conv2 (Conv2D) (None, 384, 3, 3)    332160      zero_padding2d_19[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_pool_bn (BatchNorm (None, 96, 1, 1)     384         inception_5a_pool_conv[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_1x1_conv (Conv2D)  (None, 256, 3, 3)    262400      concatenate_4[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_3x3_bn2 (BatchNorm (None, 384, 3, 3)    1536        inception_5a_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_31 (Activation)      (None, 96, 1, 1)     0           inception_5a_pool_bn[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "inception_5a_1x1_bn (BatchNorma (None, 256, 3, 3)    1024        inception_5a_1x1_conv[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_30 (Activation)      (None, 384, 3, 3)    0           inception_5a_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_20 (ZeroPadding2 (None, 96, 3, 3)     0           activation_31[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_32 (Activation)      (None, 256, 3, 3)    0           inception_5a_1x1_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_5 (Concatenate)     (None, 736, 3, 3)    0           activation_30[0][0]              \n",
            "                                                                 zero_padding2d_20[0][0]          \n",
            "                                                                 activation_32[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_3x3_conv1 (Conv2D) (None, 96, 3, 3)     70752       concatenate_5[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_3x3_bn1 (BatchNorm (None, 96, 3, 3)     384         inception_5b_3x3_conv1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_33 (Activation)      (None, 96, 3, 3)     0           inception_5b_3x3_bn1[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2D)  (None, 736, 1, 1)    0           concatenate_5[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_21 (ZeroPadding2 (None, 96, 5, 5)     0           activation_33[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_pool_conv (Conv2D) (None, 96, 1, 1)     70752       max_pooling2d_5[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_3x3_conv2 (Conv2D) (None, 384, 3, 3)    332160      zero_padding2d_21[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_pool_bn (BatchNorm (None, 96, 1, 1)     384         inception_5b_pool_conv[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_1x1_conv (Conv2D)  (None, 256, 3, 3)    188672      concatenate_5[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_3x3_bn2 (BatchNorm (None, 384, 3, 3)    1536        inception_5b_3x3_conv2[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_35 (Activation)      (None, 96, 1, 1)     0           inception_5b_pool_bn[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "inception_5b_1x1_bn (BatchNorma (None, 256, 3, 3)    1024        inception_5b_1x1_conv[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_34 (Activation)      (None, 384, 3, 3)    0           inception_5b_3x3_bn2[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_22 (ZeroPadding2 (None, 96, 3, 3)     0           activation_35[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_36 (Activation)      (None, 256, 3, 3)    0           inception_5b_1x1_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_6 (Concatenate)     (None, 736, 3, 3)    0           activation_34[0][0]              \n",
            "                                                                 zero_padding2d_22[0][0]          \n",
            "                                                                 activation_36[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_3 (AveragePoo (None, 736, 1, 1)    0           concatenate_6[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "flatten (Flatten)               (None, 736)          0           average_pooling2d_3[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "dense_layer (Dense)             (None, 128)          94336       flatten[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "lambda (Lambda)                 (None, 128)          0           dense_layer[0][0]                \n",
            "==================================================================================================\n",
            "Total params: 3,743,280\n",
            "Trainable params: 3,733,968\n",
            "Non-trainable params: 9,312\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6IQJiKtFJNIP",
        "outputId": "8c643750-6515-4aee-a5d9-d21d7531cc23",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(\"Total Params:\", FRmodel.count_params())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total Params: 3743280\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RO99N2__VYzP"
      },
      "source": [
        "...."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3_p4B3LcVYnD"
      },
      "source": [
        "..."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-xh2UDWdVYW-"
      },
      "source": [
        "...."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w0GzpLJKVYB_"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y_9fm5D_TZUG"
      },
      "source": [
        "# TRIPLET LOSS\n",
        "\n",
        "This is the loss function that will preferably be used to train the model in case the model needs to be trained at all to fit the requirement of the user"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CZyezP-RMmfH"
      },
      "source": [
        "\n",
        "def triplet_loss(y_true, y_pred, alpha = 0.2):     # alpha taken as 0.2\n",
        "  \n",
        "  anchor, positive, negative = y_pred[0], y_pred[1], y_pred[2]\n",
        "  \n",
        "  #print(anchor.shape)\n",
        "  \n",
        "  pos_dist = tf.reduce_sum(tf.square(tf.subtract(anchor, positive)), axis = -1)\n",
        "  \n",
        "  neg_dist = tf.reduce_sum(tf.square(tf.subtract(anchor, negative)), axis = -1)\n",
        "  \n",
        "  basic_loss = tf.add(tf.subtract(pos_dist, neg_dist), alpha)\n",
        "  \n",
        "  loss = tf.reduce_sum(tf.maximum(basic_loss, 0.0), axis = None)\n",
        "  \n",
        "  return loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ps69Vh5hTw6c"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uTYK6GB9TV8g"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MwmpcAiqWbGx"
      },
      "source": [
        "....."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uFI-x8QzWa50"
      },
      "source": [
        "...."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iVM_NzqRWasY"
      },
      "source": [
        "..."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MLijdEltWaiK"
      },
      "source": [
        "...."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VQFv5i5PWaXO"
      },
      "source": [
        "...."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bscaj8bLWizo"
      },
      "source": [
        "## **COMPILING THE MODEL**\n",
        "\n",
        "Only compiling model 1 (Inceptionv2) as of now"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M9axdiX6uV8P"
      },
      "source": [
        "Loading pretrained weights into the model skeleton. The weights are pretrained on the imageNet dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rXkkKT5KVf4y"
      },
      "source": [
        "FRmodel.compile(optimizer = 'adam', loss = triplet_loss, metrics = ['accuracy'])\n",
        "load_weights_from_FaceNet(FRmodel)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cK9JqQPMVoxn"
      },
      "source": [
        "# done for the day : 15.09.2020"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ugf3XTgd2P9f"
      },
      "source": [
        "import tensorflow.compat.v1 as tf\n",
        "tf.disable_v2_behavior()\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lceOY9GM6d97"
      },
      "source": [
        "with  tf.Session() as test:\n",
        "    tf.set_random_seed(1)\n",
        "    y_true = (None, None, None)\n",
        "    y_pred = (tf.random_normal([3, 128], mean=6, stddev=0.1, seed = 1),\n",
        "              tf.random_normal([3, 128], mean=1, stddev=1, seed = 1),\n",
        "              tf.random_normal([3, 128], mean=3, stddev=4, seed = 1))\n",
        "    loss = triplet_loss(y_true, y_pred)\n",
        "    \n",
        "    print(\"loss = \" + str(loss.eval()))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WvLvZdp-aixq"
      },
      "source": [
        "tf.executing_eagerly()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w4kpk6ZZuuFz"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c1K_3xdjutqs"
      },
      "source": [
        "# Creating the dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DgP0QG3J_VRV"
      },
      "source": [
        "database = {}\n",
        "database[\"samarpan\"] = img_to_encoding(\"samarpan.jpg\", FRmodel)\n",
        "database[\"sampit\"] = img_to_encoding(\"sampit.jpg\", FRmodel)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nLcx1MoIZrDy"
      },
      "source": [
        "print(len(database))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FAgHM_cxZuvA",
        "outputId": "801a0ccf-2ff4-402d-e0ee-997120e590eb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tf.executing_eagerly()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "of_bemRGjVQM"
      },
      "source": [
        "....."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AcN-UAIFjV6v"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hVSNNY4KjWZl"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rT8oiixYjW5O"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S1AyOtaljXpo"
      },
      "source": [
        "....."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r6GC5zcZiDIH"
      },
      "source": [
        "# **FACE VERIFICATION METHOD**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jYAJAECaa-Rs"
      },
      "source": [
        "def faceVerification(image_path, key, database, model):\n",
        "  \n",
        "  encoding = img_to_encoding(image_path, model)\n",
        "  \n",
        "  dist = np.linalg.norm(encoding - database[key])\n",
        "  \n",
        "  if dist < 0.7:\n",
        "    print(\"It's \" + str(key) + \", welcome in!\")\n",
        "    door_open = True\n",
        "    \n",
        "  else:\n",
        "    print(\"It's not \" + str(key) + \", please go away\")\n",
        "    door_open = False\n",
        "        \n",
        "        \n",
        "    return dist, door_open"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dobCzhhDjoxh"
      },
      "source": [
        "....."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lBFQbPatjsiV"
      },
      "source": [
        "# Face Verification Operation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0wgwTl90jIGi"
      },
      "source": [
        "from google.colab import files\n",
        "uploaded = files.upload()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0uE5jKhHi-4D"
      },
      "source": [
        "faceVerification(\"samarpanTester.jpg\", \"samarpan\", database, FRmodel)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PIrV5WKZj2Ny"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G-BaP86vj2y0"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wKtlikevj-l6"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5JD4YyMYj-HN"
      },
      "source": [
        "........"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eq1hj_w2kHBs"
      },
      "source": [
        "# **FACE RECOGNITION METHOD**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FUaAP4_pkQz4"
      },
      "source": [
        "def faceRecognition(image_path, database, model):\n",
        "  \n",
        "  encoding = img_to_encoding(image_path, model)\n",
        "  min_dist = 999\n",
        "  \n",
        "  for (name, db_enc) in database.items():\n",
        "    dist = np.linalg.norm(encoding - db_enc)\n",
        "    \n",
        "    if dist < min_dist:\n",
        "      min_dist = dist\n",
        "      identity = name\n",
        "\n",
        "    \n",
        "    \n",
        "    if min_dist > 0.7:\n",
        "        print(\"Not in the database.\")\n",
        "    else:\n",
        "        print (\"it's \" + str(identity) + \", the distance is \" + str(min_dist))\n",
        "        \n",
        "    return min_dist, identity"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DfPD0xiwVohZ"
      },
      "source": [
        "......."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8zukPFK1Vvh9"
      },
      "source": [
        "#**Face Recognition Operation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H3Pz7CLaVdUq"
      },
      "source": [
        "from google.colab import files\n",
        "uploaded = files.upload()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0COdW-ndVNjc"
      },
      "source": [
        "faceRecognition(\"meTest.jpg\", database, FRmodel)"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}