A popular method to boost the performance of a network for computer vision tasks is to add data augmentation. By using data augmentation during training time, you can increase the size of the training set. As a consequence, you make your model more robust to slight variations in the data. In Chapter 7, Computer Vision, we demonstrated some data augmentation techniques. In the following recipe, we will be using Keras and its ImageDataGenerator
for data augmentation.
- We start by importing all libraries as usual:
from keras.preprocessing.image import ImageDataGenerator from keras.models import Sequential from keras.layers import Dense, Dropout, Activation, Flatten from keras.layers import Conv2D, MaxPooling2D from keras.optimizers import Adam from keras.callbacks import EarlyStopping, TensorBoard, ModelCheckpoint from keras.datasets import cifar10
- We load and pre-process the training and validation data as follows:
(X_train, y_train...