The `ImageDataGenerator` class in the Python Keras library is used for real-time data augmentation of images during the model training process. It generates image data batches with optional augmentations, such as scaling, rotation, shearing, and flipping, to increase the diversity of training data and improve the model's generalization ability. This class also provides functionalities for data normalization, saving augmented images to disk, and handling image labels. It is a powerful tool for efficiently training image classification models by dynamically augmenting and feeding the data during the training process.
Python ImageDataGenerator - 60 examples found. These are the top rated real world Python examples of keras.preprocessing.image.ImageDataGenerator extracted from open source projects. You can rate examples to help us improve the quality of examples.