CatVsDogs.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 2,
  "metadata": {
    "accelerator": "TPU",
    "colab": {
      "name": "ProjectML.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "import os\n",
        "from google.colab import drive\n",
        "drive.mount('/gdrive', force_remount=True)\n",
        "drive_path = '/gdrive/MyDrive/Data/dataset_dogs_vs_cats.zip'\n",
        "local_path = '/content'\n",
        "\n",
        "# Copy the zip file and move it up one level (AKA out of the drive folder)\n",
        "!cp '{drive_path}' .\n",
        "\n",
        "# Navigate to the copied file and unzip it quietly\n",
        "os.chdir(local_path)\n",
        "!unzip -q 'dataset_dogs_vs_cats.zip'"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /gdrive\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uxjCFMTixgei",
        "outputId": "e69bb81f-1a1c-45c7-f833-51a56d247f90"
      }
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "IPcMHIIjip1H"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# plot cat photos from the dogs vs cats dataset\n",
        "from matplotlib import pyplot\n",
        "from matplotlib.image import imread\n",
        "# define location of dataset\n",
        "folder = 'train/'\n",
        "# plot first few images\n",
        "for i in range(9):\n",
        "\t# define subplot\n",
        "\tpyplot.subplot(330 + 1 + i)\n",
        "\t# define filename\n",
        "\tfilename = folder + 'cat.' + str(i) + '.jpg'    # cat or dog\n",
        "\t# load image pixels\n",
        "\timage = imread(filename)\n",
        "\t# plot raw pixel data\n",
        "\tpyplot.imshow(image)\n",
        "# show the figure\n",
        "pyplot.show()"
      ],
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 9 Axes>"
            ],
            "image/png": "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"
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        },
        "id": "Sfb8iZKW2Ae4",
        "outputId": "0df6bbba-0199-483f-8bba-ce9fbe0c3aa3"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# organize dataset into a useful structure\n",
        "from os import makedirs\n",
        "from os import listdir\n",
        "from shutil import copyfile\n",
        "from random import seed\n",
        "from random import random\n",
        "# create directories\n",
        "dataset_home = 'dataset_dogs_vs_cats/'\n",
        "subdirs = ['train/', 'test/']\n",
        "for subdir in subdirs:\n",
        "\t# create label subdirectories\n",
        "\tlabeldirs = ['dogs/', 'cats/']\n",
        "\tfor labldir in labeldirs:\n",
        "\t\tnewdir = dataset_home + subdir + labldir\n",
        "\t\tmakedirs(newdir, exist_ok=True)\n",
        "# seed random number generator\n",
        "seed(1)\n",
        "# define ratio of pictures to use for validation\n",
        "val_ratio = 0.25\n",
        "# copy training dataset images into subdirectories\n",
        "src_directory = 'train/'\n",
        "for file in listdir(src_directory):\n",
        "\tsrc = src_directory + '/' + file\n",
        "\tdst_dir = 'train/'\n",
        "\tif random() < val_ratio:\n",
        "\t\tdst_dir = 'test/'\n",
        "\tif file.startswith('cat'):\n",
        "\t\tdst = dataset_home + dst_dir + 'cats/'  + file\n",
        "\t\tcopyfile(src, dst)\n",
        "\telif file.startswith('dog'):\n",
        "\t\tdst = dataset_home + dst_dir + 'dogs/'  + file\n",
        "\t\tcopyfile(src, dst)"
      ],
      "outputs": [],
      "metadata": {
        "id": "dnhDYUi42oFg"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# baseline model for the dogs vs cats dataset\n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "# define cnn model\n",
        "def define_model():\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same', input_shape=(200, 200, 3)))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Flatten())\n",
        "\tmodel.add(Dense(128, activation='relu', kernel_initializer='he_uniform'))\n",
        "\tmodel.add(Dense(1, activation='sigmoid'))\n",
        "\t# compile model\n",
        "\topt = SGD(lr=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\treturn model\n",
        "\n",
        "def summarize_diagnostics(history):\n",
        "    # plot loss\n",
        "    pyplot.subplot(211)\n",
        "    pyplot.title('Cross Entropy Loss')\n",
        "    pyplot.plot(history.history['loss'], color='red', label='train')\n",
        "    pyplot.plot(history.history['val_loss'], color='black', label='test')\n",
        "    pyplot.tight_layout()\n",
        "    # plot accuracy\n",
        "    pyplot.subplot(212)\n",
        "    pyplot.title('Classification Accuracy')\n",
        "    pyplot.plot(history.history['accuracy'], color='red', label='train')\n",
        "    pyplot.plot(history.history['val_accuracy'], color='black', label= 'test')\n",
        "    pyplot.tight_layout()\n",
        "    # save plot to file\n",
        "    filename = sys.argv[0].split('/')[-1]\n",
        "    pyplot.savefig(filename + '_plot.png')\n",
        "    pyplot.close()\n",
        "\n",
        "# run the test harness for evaluating a model\n",
        "def run_test_harness():\n",
        "\t# define model\n",
        "\tmodel = define_model()\n",
        "\t# create data generator\n",
        "\tdatagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\t# prepare iterators\n",
        "\ttrain_it = datagen.flow_from_directory('dataset_dogs_vs_cats/train/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\ttest_it = datagen.flow_from_directory('dataset_dogs_vs_cats/test/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\t# fit model\n",
        "\thistory = model.fit(train_it, steps_per_epoch=len(train_it),\n",
        "\t\tvalidation_data=test_it, validation_steps=len(test_it), epochs=20, verbose=0)\n",
        "\t# evaluate model\n",
        "\t_, acc = model.evaluate(test_it, steps=len(test_it), verbose=0)\n",
        "\tprint('> %.3f' % (acc * 100.0))\n",
        "\t# learning curves\n",
        "\tsummarize_diagnostics(history)\n",
        "\n",
        "# entry point, run the test harness\n",
        "run_test_harness()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 18697 images belonging to 2 classes.\n",
            "Found 6303 images belonging to 2 classes.\n",
            "> 71.220\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7DSIHoAu4TJA",
        "outputId": "d3556705-5bc4-42f9-83e1-fa4c88c5fd1e"
      }
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "zh4gBZMhRIn5"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# Two block VGG model for the dogs vs cats dataset\n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "# define cnn model\n",
        "def define_model():\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same', input_shape=(200, 200, 3)))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))   \n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\n",
        "\tmodel.add(Flatten())\n",
        "\tmodel.add(Dense(128, activation='relu', kernel_initializer='he_uniform'))\n",
        "\tmodel.add(Dense(1, activation='sigmoid'))\n",
        "\t# compile model\n",
        "\topt = SGD(lr=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\treturn model\n",
        "\n",
        "def summarize_diagnostics(history):\n",
        "    # plot loss\n",
        "    pyplot.subplot(211)\n",
        "    pyplot.title('Cross Entropy Loss')\n",
        "    pyplot.plot(history.history['loss'], color='red', label='train')\n",
        "    pyplot.plot(history.history['val_loss'], color='black', label='test')\n",
        "    pyplot.tight_layout()\n",
        "    # plot accuracy\n",
        "    pyplot.subplot(212)\n",
        "    pyplot.title('Classification Accuracy')\n",
        "    pyplot.plot(history.history['accuracy'], color='red', label='train')\n",
        "    pyplot.plot(history.history['val_accuracy'], color='black', label= 'test')\n",
        "    pyplot.tight_layout()\n",
        "    # save plot to file\n",
        "    filename = sys.argv[0].split('/')[-1]\n",
        "    pyplot.savefig(filename + '_plot.png')\n",
        "    pyplot.close()\n",
        "\n",
        "# run the test harness for evaluating a model\n",
        "def run_test_harness():\n",
        "\t# define model\n",
        "\tmodel = define_model()\n",
        "\t# create data generator\n",
        "\tdatagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\t# prepare iterators\n",
        "\ttrain_it = datagen.flow_from_directory('dataset_dogs_vs_cats/train/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\ttest_it = datagen.flow_from_directory('dataset_dogs_vs_cats/test/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\t# fit model\n",
        "\thistory = model.fit(train_it, steps_per_epoch=len(train_it),\n",
        "\t\tvalidation_data=test_it, validation_steps=len(test_it), epochs=20, verbose=0)\n",
        "\t# evaluate model\n",
        "\t_, acc = model.evaluate(test_it, steps=len(test_it), verbose=0)\n",
        "\tprint('> %.3f' % (acc * 100.0))\n",
        "\t# learning curves\n",
        "\tsummarize_diagnostics(history)\n",
        "\n",
        "# entry point, run the test harness\n",
        "run_test_harness()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 18697 images belonging to 2 classes.\n",
            "Found 6303 images belonging to 2 classes.\n",
            "> 75.472\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8Y0N_Pp6GRah",
        "outputId": "cae9e41d-479d-430b-c6ae-22ae8bb09ac0"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# Three block VGG model for the dogs vs cats dataset\n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "# define cnn model\n",
        "def define_model():\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same', input_shape=(200, 200, 3)))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))   \n",
        "\tmodel.add(MaxPooling2D((2, 2)))\t\n",
        "\tmodel.add(Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))   \n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\n",
        "\tmodel.add(Flatten())\n",
        "\tmodel.add(Dense(128, activation='relu', kernel_initializer='he_uniform'))\n",
        "\tmodel.add(Dense(1, activation='sigmoid'))\n",
        "\t# compile model\n",
        "\topt = SGD(lr=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\treturn model\n",
        "\n",
        "def summarize_diagnostics(history):\n",
        "    # plot loss\n",
        "    pyplot.subplot(211)\n",
        "    pyplot.title('Cross Entropy Loss')\n",
        "    pyplot.plot(history.history['loss'], color='red', label='train')\n",
        "    pyplot.plot(history.history['val_loss'], color='black', label='test')\n",
        "    pyplot.tight_layout()\n",
        "    # plot accuracy\n",
        "    pyplot.subplot(212)\n",
        "    pyplot.title('Classification Accuracy')\n",
        "    pyplot.plot(history.history['accuracy'], color='red', label='train')\n",
        "    pyplot.plot(history.history['val_accuracy'], color='black', label= 'test')\n",
        "    pyplot.tight_layout()\n",
        "    # save plot to file\n",
        "    filename = sys.argv[0].split('/')[-1]\n",
        "    pyplot.savefig(filename + '_plot.png')\n",
        "    pyplot.close()\n",
        "\n",
        "# run the test harness for evaluating a model\n",
        "def run_test_harness():\n",
        "\t# define model\n",
        "\tmodel = define_model()\n",
        "\t# create data generator\n",
        "\tdatagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\t# prepare iterators\n",
        "\ttrain_it = datagen.flow_from_directory('dataset_dogs_vs_cats/train/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\ttest_it = datagen.flow_from_directory('dataset_dogs_vs_cats/test/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\t# fit model\n",
        "\thistory = model.fit(train_it, steps_per_epoch=len(train_it),\n",
        "\t\tvalidation_data=test_it, validation_steps=len(test_it), epochs=20, verbose=0)\n",
        "\t# evaluate model\n",
        "\t_, acc = model.evaluate(test_it, steps=len(test_it), verbose=0)\n",
        "\tprint('> %.3f' % (acc * 100.0))\n",
        "\t# learning curves\n",
        "\tsummarize_diagnostics(history)\n",
        "\n",
        "# entry point, run the test harness\n",
        "run_test_harness()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 18697 images belonging to 2 classes.\n",
            "Found 6303 images belonging to 2 classes.\n",
            "> 78.471\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zkgK2vk8L55C",
        "outputId": "f37b8678-6294-42ce-a8e6-32830ecdc7bf"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# Three block VGG model with data augmentation for the dogs vs cats dataset\n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "# define cnn model\n",
        "def define_model():\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same', input_shape=(200, 200, 3)))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Flatten())\n",
        "\tmodel.add(Dense(128, activation='relu', kernel_initializer='he_uniform'))\n",
        "\tmodel.add(Dense(1, activation='sigmoid'))\n",
        "\t# compile model\n",
        "\topt = SGD(lr=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\treturn model\n",
        "\n",
        "# plot diagnostic learning curves\n",
        "def summarize_diagnostics(history):\n",
        "\t# plot loss\n",
        "\tpyplot.subplot(211)\n",
        "\tpyplot.title('Cross Entropy Loss')\n",
        "\tpyplot.plot(history.history['loss'], color='red', label='train')\n",
        "\tpyplot.plot(history.history['val_loss'], color='black', label='test')\n",
        "\t# plot accuracy\n",
        "\tpyplot.subplot(212)\n",
        "\tpyplot.title('Classification Accuracy')\n",
        "\tpyplot.plot(history.history['accuracy'], color='red', label='train')\n",
        "\tpyplot.plot(history.history['val_accuracy'], color='black', label='test')\n",
        "\t# save plot to file\n",
        "\tfilename = sys.argv[0].split('/')[-1]\n",
        "\tpyplot.savefig(filename + '_plot.png')\n",
        "\tpyplot.close()\n",
        "\n",
        "# run the test harness for evaluating a model\n",
        "def run_test_harness():\n",
        "\t# define model\n",
        "\tmodel = define_model()\n",
        "\t# create data generators\n",
        "\ttrain_datagen = ImageDataGenerator(rescale=1.0/255.0,\n",
        "\t\twidth_shift_range=0.1, height_shift_range=0.1, horizontal_flip=True)\n",
        "\ttest_datagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\t# prepare iterators\n",
        "\ttrain_it = train_datagen.flow_from_directory('dataset_dogs_vs_cats/train/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\ttest_it = test_datagen.flow_from_directory('dataset_dogs_vs_cats/test/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(200, 200))\n",
        "\t# fit model\n",
        "\thistory = model.fit(train_it, steps_per_epoch=len(train_it),\n",
        "\t\tvalidation_data=test_it, validation_steps=len(test_it), epochs=50, verbose=0)\n",
        "\t# evaluate model\n",
        "\t_, acc = model.evaluate(test_it, steps=len(test_it), verbose=0)\n",
        "\tprint('> %.3f' % (acc * 100.0))\n",
        "\t# learning curves\n",
        "\tsummarize_diagnostics(history)\n",
        "\n",
        "# entry point, run the test harness\n",
        "run_test_harness()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 18697 images belonging to 2 classes.\n",
            "Found 6303 images belonging to 2 classes.\n",
            "> 86.625\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WpS_2nYYT_6v",
        "outputId": "9bf0dbc4-1305-41d6-ed0d-aec6a15f824d"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# vgg16 model used for transfer learning on the dogs and cats dataset\n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from keras.utils import to_categorical\n",
        "from keras.applications.vgg16 import VGG16\n",
        "from keras.models import Model\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "# define cnn model\n",
        "def define_model():\n",
        "\t# load model\n",
        "\tmodel = VGG16(include_top=False, input_shape=(224, 224, 3))\n",
        "\t# mark loaded layers as not trainable\n",
        "\tfor layer in model.layers:\n",
        "\t\tlayer.trainable = False\n",
        "\t# add new classifier layers\n",
        "\tflat1 = Flatten()(model.layers[-1].output)\n",
        "\tclass1 = Dense(128, activation='relu', kernel_initializer='he_uniform')(flat1)\n",
        "\toutput = Dense(1, activation='sigmoid')(class1)\n",
        "\t# define new model\n",
        "\tmodel = Model(inputs=model.inputs, outputs=output)\n",
        "\t# compile model\n",
        "\topt = SGD(lr=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\treturn model\n",
        "\n",
        "# plot diagnostic learning curves\n",
        "def summarize_diagnostics(history):\n",
        "\t# plot loss\n",
        "\tpyplot.subplot(211)\n",
        "\tpyplot.title('Cross Entropy Loss')\n",
        "\tpyplot.plot(history.history['loss'], color='red', label='train')\n",
        "\tpyplot.plot(history.history['val_loss'], color='black', label='test')\n",
        "\tpyplot.tight_layout()\n",
        "\t# plot accuracy\n",
        "\tpyplot.subplot(212)\n",
        "\tpyplot.title('Classification Accuracy')\n",
        "\tpyplot.plot(history.history['accuracy'], color='red', label='train')\n",
        "\tpyplot.plot(history.history['val_accuracy'], color='black', label='test')\n",
        "\tpyplot.tight_layout()\n",
        "\t# save plot to file\n",
        "\tfilename = sys.argv[0].split('/')[-1]\n",
        "\tpyplot.savefig(filename + '_plot.png')\n",
        "\tpyplot.close()\n",
        "\n",
        "# run the test harness for evaluating a model\n",
        "def run_test_harness():\n",
        "\t# define model\n",
        "\tmodel = define_model()\n",
        "\t# create data generator\n",
        "\tdatagen = ImageDataGenerator(featurewise_center=True)\n",
        "\t# specify imagenet mean values for centering\n",
        "\tdatagen.mean = [123.68, 116.779, 103.939]\n",
        "\t# prepare iterator\n",
        "\ttrain_it = datagen.flow_from_directory('dataset_dogs_vs_cats/train/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(224, 224))\n",
        "\ttest_it = datagen.flow_from_directory('dataset_dogs_vs_cats/test/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(224, 224))\n",
        "\t# fit model\n",
        "\thistory = model.fit(train_it, steps_per_epoch=len(train_it),\n",
        "\t\tvalidation_data=test_it, validation_steps=len(test_it), epochs=10, verbose=1)\n",
        "\t# evaluate model\n",
        "\t_, acc = model.evaluate(test_it, steps=len(test_it), verbose=0)\n",
        "\tprint('> %.3f' % (acc * 100.0))\n",
        "\t# learning curves\n",
        "\tsummarize_diagnostics(history)\n",
        " \n",
        "# entry point, run the test harness\n",
        "run_test_harness()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 18697 images belonging to 2 classes.\n",
            "Found 6303 images belonging to 2 classes.\n",
            "Epoch 1/10\n",
            "293/293 [==============================] - 186s 631ms/step - loss: 1.0733 - accuracy: 0.9347 - val_loss: 0.0740 - val_accuracy: 0.9757\n",
            "Epoch 2/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 0.0309 - accuracy: 0.9890 - val_loss: 0.0879 - val_accuracy: 0.9776\n",
            "Epoch 3/10\n",
            "293/293 [==============================] - 185s 630ms/step - loss: 0.0102 - accuracy: 0.9977 - val_loss: 0.0992 - val_accuracy: 0.9767\n",
            "Epoch 4/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 0.0035 - accuracy: 0.9996 - val_loss: 0.1142 - val_accuracy: 0.9775\n",
            "Epoch 5/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 0.0014 - accuracy: 0.9999 - val_loss: 0.1221 - val_accuracy: 0.9767\n",
            "Epoch 6/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 0.0011 - accuracy: 1.0000 - val_loss: 0.1312 - val_accuracy: 0.9768\n",
            "Epoch 7/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 7.5303e-04 - accuracy: 0.9999 - val_loss: 0.1364 - val_accuracy: 0.9764\n",
            "Epoch 8/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 4.4686e-04 - accuracy: 0.9999 - val_loss: 0.1400 - val_accuracy: 0.9767\n",
            "Epoch 9/10\n",
            "293/293 [==============================] - 184s 629ms/step - loss: 5.6103e-04 - accuracy: 0.9997 - val_loss: 0.1431 - val_accuracy: 0.9764\n",
            "Epoch 10/10\n",
            "293/293 [==============================] - 183s 626ms/step - loss: 3.7819e-04 - accuracy: 0.9998 - val_loss: 0.1461 - val_accuracy: 0.9767\n",
            "> 97.668\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8cArkRZ3lN5-",
        "outputId": "936b9195-dc67-4547-9bf8-a179fdcf288a"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# organize dataset into a useful structure\n",
        "from os import makedirs\n",
        "from os import listdir\n",
        "from shutil import copyfile\n",
        "# create directories\n",
        "dataset_home = 'finalize_dogs_vs_cats/'\n",
        "# create label subdirectories\n",
        "labeldirs = ['dogs/', 'cats/']\n",
        "for labldir in labeldirs:\n",
        "\tnewdir = dataset_home + labldir\n",
        "\tmakedirs(newdir, exist_ok=True)\n",
        "# copy training dataset images into subdirectories\n",
        "src_directory = 'train/'\n",
        "for file in listdir(src_directory):\n",
        "\tsrc = src_directory + '/' + file\n",
        "\tif file.startswith('cat'):\n",
        "\t\tdst = dataset_home + 'cats/'  + file\n",
        "\t\tcopyfile(src, dst)\n",
        "\telif file.startswith('dog'):\n",
        "\t\tdst = dataset_home + 'dogs/'  + file\n",
        "\t\tcopyfile(src, dst)"
      ],
      "outputs": [],
      "metadata": {
        "id": "N49HVTZyvmLS"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# saving the final model VGG16 model with custom classified classifier to file\n",
        "from keras.applications.vgg16 import VGG16\n",
        "from keras.models import Model\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        " \n",
        "# define cnn model\n",
        "def define_model():\n",
        "\t# load model\n",
        "\tmodel = VGG16(include_top=False, input_shape=(224, 224, 3))\n",
        "\t# mark loaded layers as not trainable\n",
        "\tfor layer in model.layers:\n",
        "\t\tlayer.trainable = False\n",
        "\t# add new classifier layers\n",
        "\tflat1 = Flatten()(model.layers[-1].output)\n",
        "\tclass1 = Dense(128, activation='relu', kernel_initializer='he_uniform')(flat1)\n",
        "\toutput = Dense(1, activation='sigmoid')(class1)\n",
        "\t# define new model\n",
        "\tmodel = Model(inputs=model.inputs, outputs=output)\n",
        "\t# compile model\n",
        "\topt = SGD(lr=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\treturn model\n",
        " \n",
        "# run the test harness for evaluating a model\n",
        "def run_test_harness():\n",
        "# define model\n",
        "\tmodel = define_model()\n",
        "\t# create data generator\n",
        "\tdatagen = ImageDataGenerator(featurewise_center=True)\n",
        "\t# specify imagenet mean values for centering\n",
        "\tdatagen.mean = [123.68, 116.779, 103.939]\n",
        "\t# prepare iterator\n",
        "\ttrain_it = datagen.flow_from_directory('finalize_dogs_vs_cats/',\n",
        "\t\tclass_mode='binary', batch_size=64, target_size=(224, 224))\n",
        "\t# fit model\n",
        "\tmodel.fit(train_it, steps_per_epoch=len(train_it), epochs=10, verbose=0)\n",
        "\t# save model\n",
        "\tmodel.save('final_model.h5')\n",
        " \n",
        "# entry point, run the test harness\n",
        "run_test_harness()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "58892288/58889256 [==============================] - 0s 0us/step\n",
            "Found 25000 images belonging to 2 classes.\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wZ_oE524W1T7",
        "outputId": "dd8be336-8c56-48d9-a95c-bfd00490559f"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "source": [
        "# make a prediciton for a new image\n",
        "from keras.preprocessing.image import load_img\n",
        "from keras.preprocessing.image import img_to_array\n",
        "from keras.models import load_model\n",
        "\n",
        "# load and prepare the image\n",
        "def load_image(filename):\n",
        "    # load the image\n",
        "    img = load_img(filename, target_size=(224,224))\n",
        "    # convert to array\n",
        "    img = img_to_array(img)\n",
        "    # reshape into a single sample with 3 channels\n",
        "    img = img.reshape(1,224,224,3)\n",
        "    # center pixel data\n",
        "    img = img.astype('float32')\n",
        "    img = img - [123.68, 116.779, 103.939]\n",
        "    return img\n",
        "\n",
        "#load an image and predict the class\n",
        "def run_example():\n",
        "    # load the image\n",
        "    img = load_image('sample_image.jpg')\n",
        "    #load model\n",
        "    model = load_model('final_model.h5')\n",
        "    # predict the class\n",
        "    result = model.predict(img)\n",
        "    print(result[0])\n",
        "\n",
        "    \n",
        "\n",
        "# entry point, run the example\n",
        "run_example()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[1.]\n"
          ]
        }
      ],
      "metadata": {
        "id": "h9OKuFZL9a92",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5ef79d1f-fd68-4054-bb65-85ed2a965eb3"
      }
    }
  ]
}