{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.4"
},
"colab": {
"name": "classify_images_cats-vs-dogs.ipynb",
"provenance": [],
"collapsed_sections": []
},
"accelerator": "GPU"
},
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "aZ6fXTzdYqbS",
"colab_type": "text"
},
"source": [
"# CS568:Deep Learning Spring 2020 "
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "UYIHpg--Yqbc",
"colab_type": "text"
},
"source": [
"Mount your Google drive by using this code snippet"
]
},
{
"cell_type": "code",
"metadata": {
"id": "FunfJyntYqbj",
"colab_type": "code",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 34
},
"outputId": "079facca-3a20-4435-f89c-9c78351f9af9"
},
"source": [
"from google.colab import drive\n",
"drive.mount('/content/drive')"
],
"execution_count": 2,
"outputs": [
{
"output_type": "stream",
"text": [
"Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "tUQJxmN5Yqb3",
"colab_type": "text"
},
"source": [
"In this exercise, we will build a neural network model from scratch that is able to distinguish dogs from cats. \n",
"\n",
"**NOTE:** The 2200 images used in this exercise are excerpted from the [\"Dogs vs. Cats\"](https://www.kaggle.com/c/dogs-vs-cats/data) dataset available on Kaggle, which contains 25,000 images. We use a subset of the full dataset here.\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "GJenTzZTYqb-",
"colab_type": "text"
},
"source": [
"#### 1.1 Load dataset\n",
" "
]
},
{
"cell_type": "code",
"metadata": {
"id": "bo8kR9KXYqcH",
"colab_type": "code",
"outputId": "cf79ef14-cd08-49d9-c052-56cb27937327",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 68
}
},
"source": [
"from keras.preprocessing.image import ImageDataGenerator\n",
"\n",
"img_height, img_width = 128, 128\n",
"\n",
"train_data_dir = '/content/drive/My Drive/CS568-DeepLearning-(Spring2020)/dataset/dogs-vs-cats/train/'\n",
"validation_data_dir = '/content/drive/My Drive/CS568-DeepLearning-(Spring2020)/dataset/dogs-vs-cats/val/' \n",
"test_data_dir = '/content/drive/My Drive/CS568-DeepLearning-(Spring2020)/dataset/dogs-vs-cats/'\n",
"\n",
"# normalize the pixel values from [0, 255] to [0, 1] interval\n",
"datagen = ImageDataGenerator(rescale=1./255)\n",
"\n",
"# automagically retrieve images and their classes for train and validation sets\n",
"batch_size = 32 # e.g 4, 8, 16, 32, 64\n",
"\n",
"train_generator = datagen.flow_from_directory(\n",
" train_data_dir,\n",
" target_size=(img_width, img_height),\n",
" color_mode = 'grayscale',\n",
" batch_size=batch_size,\n",
" class_mode='binary')\n",
"\n",
"validation_generator = datagen.flow_from_directory(\n",
" validation_data_dir,\n",
" target_size=(img_width, img_height),\n",
" color_mode = 'grayscale',\n",
" batch_size=batch_size,\n",
" class_mode='binary')\n",
"\n",
"test_generator = datagen.flow_from_directory(\n",
" test_data_dir,\n",
" target_size=(img_width, img_height),\n",
" color_mode = 'grayscale',\n",
" classes = ['test'],\n",
" class_mode=None)"
],
"execution_count": 9,
"outputs": [
{
"output_type": "stream",
"text": [
"Found 2067 images belonging to 2 classes.\n",
"Found 217 images belonging to 2 classes.\n",
"Found 100 images belonging to 1 classes.\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "Og2by6k6Yqcd",
"colab_type": "code",
"outputId": "72585b0f-45c9-4966-f6c4-a0fecc432233",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 1000
}
},
"source": [
"from keras.models import Sequential\n",
"from keras import layers\n",
"from keras.layers import Dropout, Flatten, Dense\n",
"from keras.optimizers import SGD\n",
"\n",
"img_width, img_height = 128, 128\n",
"input_dim = 128*128\n",
"output_dim = 1\n",
"\n",
"def lr():\n",
" # initialize model\n",
" model = Sequential()\n",
" # flatten or vectorize input layer images\n",
" model.add(Flatten(input_shape=(img_height,img_width,1)))\n",
" # add an input layer and output layer with activation function\n",
" model.add(Dense(output_dim, input_dim=input_dim, activation='sigmoid')) \n",
" return model\n",
"\n",
"# define model\n",
"model = lr()\n",
"# define optimizer\n",
"sgd = SGD(lr=0.001)\n",
"model.compile(optimizer=sgd, loss='mse', metrics=['accuracy'])\n",
"# print model information\n",
"model.summary()\n",
"\n",
"\n",
"nb_epoch = 100\n",
"nb_train_samples = 2000\n",
"nb_validation_samples = 200\n",
"\n",
"import time\n",
"start = time.time()\n",
"model_info = model.fit_generator(\n",
" train_generator,\n",
" samples_per_epoch=nb_train_samples,\n",
" nb_epoch=nb_epoch,\n",
" validation_data=validation_generator,\n",
" nb_val_samples=nb_validation_samples)\n",
"end = time.time()\n",
"print(\"Model took %0.2f seconds to train\"%(end - start))"
],
"execution_count": 12,
"outputs": [
{
"output_type": "stream",
"text": [
"Model: \"sequential_7\"\n",
"_________________________________________________________________\n",
"Layer (type) Output Shape Param # \n",
"=================================================================\n",
"flatten_7 (Flatten) (None, 16384) 0 \n",
"_________________________________________________________________\n",
"dense_11 (Dense) (None, 1) 16385 \n",
"=================================================================\n",
"Total params: 16,385\n",
"Trainable params: 16,385\n",
"Non-trainable params: 0\n",
"_________________________________________________________________\n",
"Epoch 1/100\n"
],
"name": "stdout"
},
{
"output_type": "stream",
"text": [
"/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:39: UserWarning: The semantics of the Keras 2 argument `steps_per_epoch` is not the same as the Keras 1 argument `samples_per_epoch`. `steps_per_epoch` is the number of batches to draw from the generator at each epoch. Basically steps_per_epoch = samples_per_epoch/batch_size. Similarly `nb_val_samples`->`validation_steps` and `val_samples`->`steps` arguments have changed. Update your method calls accordingly.\n",
"/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:39: UserWarning: Update your `fit_generator` call to the Keras 2 API: `fit_generator("
]
},
"metadata": {
"tags": []
}
}
]
}
]
}