Image Augmentation

http://en.diveintodeeplearning.org/chapter_computer-vision/image-augmentation.html

Is there a way to apply custom augmentations? As an example, if I want to randomly add Salt & Pepper noise in my dataset, how should I do that?