View on TensorFlow.org
|
Run in Google Colab
|
View source on GitHub
|
 Download notebook
|
This tutorial demonstrates training a simple Convolutional Neural Network (CNN) to classify CIFAR images. Because this tutorial uses the Keras Sequential API, creating and training your model will take just a few lines of code.
Import TensorFlow
import tensorflow as tf
from tensorflow.keras import datasets, layers, models
import matplotlib.pyplot as plt
Download and prepare the CIFAR10 dataset
The CIFAR10 dataset contains 60,000 color images in 10 classes, with 6,000 images in each class. The dataset is divided into 50,000 training images and 10,000 testing images. The classes are mutually exclusive and there is no overlap between them.
(train_images, train_labels), (test_images, test_labels) = datasets.cifar10.load_data()
# Normalize pixel values to be between 0 and 1
train_images, test_images = train_images / 255.0, test_images / 255.0
Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz 170498071/170498071 [==============================] - 4s 0us/step
Verify the data
To verify that the dataset looks correct, let's plot the first 25 images from the training set and display the class name below each image:
class_names = ['airplane', 'automobile', 'bird', 'cat', 'deer',
'dog', 'frog', 'horse', 'ship', 'truck']
plt.figure(figsize=(10,10))
for i in range(25):
plt.subplot(5,5,i+1)
plt.xticks([])
plt.yticks([])
plt.grid(False)
plt.imshow(train_images[i])
# The CIFAR labels happen to be arrays,
# which is why you need the extra index
plt.xlabel(class_names[train_labels[i][0]])
plt.show()
Create the convolutional base
The 6 lines of code below define the convolutional base using a common pattern: a stack of Conv2D and MaxPooling2D layers.
As input, a CNN takes tensors of shape (image_height, image_width, color_channels), ignoring the batch size. If you are new to these dimensions, color_channels refers to (R,G,B). In this example, you will configure your CNN to process inputs of shape (32, 32, 3), which is the format of CIFAR images. You can do this by passing the argument input_shape to your first layer.
model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3)))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
Let's display the architecture of your model so far:
model.summary()
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d (Conv2D) (None, 30, 30, 32) 896
max_pooling2d (MaxPooling2 (None, 15, 15, 32) 0
D)
conv2d_1 (Conv2D) (None, 13, 13, 64) 18496
max_pooling2d_1 (MaxPoolin (None, 6, 6, 64) 0
g2D)
conv2d_2 (Conv2D) (None, 4, 4, 64) 36928
=================================================================
Total params: 56320 (220.00 KB)
Trainable params: 56320 (220.00 KB)
Non-trainable params: 0 (0.00 Byte)
_________________________________________________________________
Above, you can see that the output of every Conv2D and MaxPooling2D layer is a 3D tensor of shape (height, width, channels). The width and height dimensions tend to shrink as you go deeper in the network. The number of output channels for each Conv2D layer is controlled by the first argument (e.g., 32 or 64). Typically, as the width and height shrink, you can afford (computationally) to add more output channels in each Conv2D layer.
Add Dense layers on top
To complete the model, you will feed the last output tensor from the convolutional base (of shape (4, 4, 64)) into one or more Dense layers to perform classification. Dense layers take vectors as input (which are 1D), while the current output is a 3D tensor. First, you will flatten (or unroll) the 3D output to 1D, then add one or more Dense layers on top. CIFAR has 10 output classes, so you use a final Dense layer with 10 outputs.
model.add(layers.Flatten())
model.add(layers.Dense(64, activation='relu'))
model.add(layers.Dense(10))
Here's the complete architecture of your model:
model.summary()
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d (Conv2D) (None, 30, 30, 32) 896
max_pooling2d (MaxPooling2 (None, 15, 15, 32) 0
D)
conv2d_1 (Conv2D) (None, 13, 13, 64) 18496
max_pooling2d_1 (MaxPoolin (None, 6, 6, 64) 0
g2D)
conv2d_2 (Conv2D) (None, 4, 4, 64) 36928
flatten (Flatten) (None, 1024) 0
dense (Dense) (None, 64) 65600
dense_1 (Dense) (None, 10) 650
=================================================================
Total params: 122570 (478.79 KB)
Trainable params: 122570 (478.79 KB)
Non-trainable params: 0 (0.00 Byte)
_________________________________________________________________
The network summary shows that (4, 4, 64) outputs were flattened into vectors of shape (1024) before going through two Dense layers.
Compile and train the model
model.compile(optimizer='adam',
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])
history = model.fit(train_images, train_labels, epochs=10,
validation_data=(test_images, test_labels))
Epoch 1/10 1563/1563 [==============================] - 10s 4ms/step - loss: 1.5733 - accuracy: 0.4257 - val_loss: 1.2938 - val_accuracy: 0.5405 Epoch 2/10 1563/1563 [==============================] - 6s 4ms/step - loss: 1.1916 - accuracy: 0.5761 - val_loss: 1.1120 - val_accuracy: 0.6029 Epoch 3/10 1563/1563 [==============================] - 6s 4ms/step - loss: 1.0424 - accuracy: 0.6315 - val_loss: 1.0490 - val_accuracy: 0.6332 Epoch 4/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.9586 - accuracy: 0.6631 - val_loss: 0.9473 - val_accuracy: 0.6711 Epoch 5/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.8903 - accuracy: 0.6875 - val_loss: 0.9499 - val_accuracy: 0.6693 Epoch 6/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.8352 - accuracy: 0.7077 - val_loss: 0.9962 - val_accuracy: 0.6548 Epoch 7/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.7903 - accuracy: 0.7219 - val_loss: 0.9115 - val_accuracy: 0.6910 Epoch 8/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.7500 - accuracy: 0.7349 - val_loss: 0.8694 - val_accuracy: 0.6984 Epoch 9/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.7134 - accuracy: 0.7494 - val_loss: 0.8856 - val_accuracy: 0.6930 Epoch 10/10 1563/1563 [==============================] - 6s 4ms/step - loss: 0.6781 - accuracy: 0.7627 - val_loss: 0.8593 - val_accuracy: 0.7105
Evaluate the model
plt.plot(history.history['accuracy'], label='accuracy')
plt.plot(history.history['val_accuracy'], label = 'val_accuracy')
plt.xlabel('Epoch')
plt.ylabel('Accuracy')
plt.ylim([0.5, 1])
plt.legend(loc='lower right')
test_loss, test_acc = model.evaluate(test_images, test_labels, verbose=2)
313/313 - 1s - loss: 0.8593 - accuracy: 0.7105 - 558ms/epoch - 2ms/step
print(test_acc)
0.7105000019073486
Your simple CNN has achieved a test accuracy of over 70%. Not bad for a few lines of code! For another CNN style, check out the TensorFlow 2 quickstart for experts example that uses the Keras subclassing API and tf.GradientTape.