You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
197 lines
5.1 KiB
197 lines
5.1 KiB
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "2eaba66b",
|
|
"metadata": {},
|
|
"source": [
|
|
"Read and Display Horse or Human machine learning dataset"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "f1e48ac0",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"import tensorflow as tf\n",
|
|
"import numpy as np\n",
|
|
"import tensorflow_datasets as tfds\n",
|
|
"from tensorflow.keras import regularizers\n",
|
|
"import matplotlib.pyplot as plt"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "feda024e",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Load the horse or human dataset\n",
|
|
"#(300, 300, 3) unint8\n",
|
|
"dataset, label = tfds.load('horses_or_humans', with_info=True)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "35991dec",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Extract the horse/human class\n",
|
|
"horse_ds = dataset['train'].filter(lambda x: x['label'] == 0)\n",
|
|
"human_ds = dataset['train'].filter(lambda x: x['label'] == 1)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "fab03aa8",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Take a few examples < 16\n",
|
|
"n_examples = 5\n",
|
|
"horse_examples = horse_ds.take(n_examples)\n",
|
|
"human_examples = human_ds.take(n_examples)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "c33f1acd",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Display the examples\n",
|
|
"fig, axes = plt.subplots(1, n_examples, figsize=(15, 15))\n",
|
|
"for i, example in enumerate(human_examples):\n",
|
|
" image = example['image']\n",
|
|
" axes[i].imshow(image)\n",
|
|
" axes[i].set_title(f\"humans {i+1}\")\n",
|
|
"plt.show()\n",
|
|
"\n",
|
|
"fig, axes = plt.subplots(1, n_examples, figsize=(15, 15))\n",
|
|
"for i, example in enumerate(horse_examples):\n",
|
|
" image = example['image']\n",
|
|
" axes[i].imshow(image)\n",
|
|
" axes[i].set_title(f\"horses {i+1}\")\n",
|
|
"plt.show()\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "25f3eeb3",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Split the dataset into training and validation sets\n",
|
|
"# as_supervised: Specifies whether to return the dataset as a tuple\n",
|
|
"# of (input, label) pairs.\n",
|
|
"train_dataset, valid_dataset = tfds.load('horses_or_humans', split=['train','test'], as_supervised=True)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "29dc0e62",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Get the number of elements in the training and validation dataset\n",
|
|
"train_size = tf.data.experimental.cardinality(train_dataset).numpy()\n",
|
|
"valid_size = tf.data.experimental.cardinality(valid_dataset).numpy()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "db8aaf91",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"IMG_SIZE = 300\n",
|
|
"NUM_CLASSES = 2\n",
|
|
"\n",
|
|
"def preprocess(image, label):\n",
|
|
" image = tf.cast(image, tf.float32)\n",
|
|
"# # Resize the images to a fixed size\n",
|
|
" image = tf.image.resize(image, (IMG_SIZE, IMG_SIZE))\n",
|
|
"# # Rescale the pixel values to be between 0 and 1\n",
|
|
" image = image / 255.0\n",
|
|
" label = tf.one_hot(label, NUM_CLASSES)\n",
|
|
" return image, label"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "d59661c3",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Apply the preprocessing function to the datasets\n",
|
|
"train_dataset = train_dataset.map(preprocess)\n",
|
|
"valid_dataset = valid_dataset.map(preprocess)\n",
|
|
"\n",
|
|
"# Batch and shuffle the datasets\n",
|
|
"train_dataset = train_dataset.shuffle(2000).batch(80)\n",
|
|
"valid_dataset = valid_dataset.batch(20)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "9399bc99",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Get the number of elements in the trainingand validation dataset\n",
|
|
"train_size = tf.data.experimental.cardinality(train_dataset).numpy()\n",
|
|
"valid_size = tf.data.experimental.cardinality(valid_dataset).numpy()\n",
|
|
"print(\"Training dataset size:\", train_size)\n",
|
|
"print(\"Validation dataset size:\", valid_size)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "13af7d53",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# Store images and labels of the validation data for predictions\n",
|
|
"for images, labels in valid_dataset:\n",
|
|
" x_val = images\n",
|
|
" y_val = labels\n",
|
|
" \n",
|
|
"print(x_val.shape, y_val.shape)"
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3 (ipykernel)",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.8.16"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 5
|
|
}
|