This library performs image augmentation for machine learning experiments, converting a set of input images into a much larger set of slightly altered variants. It supports augmenting images along with associated annotations including bounding boxes, keypoints, polygons, segmentation maps, and heatmaps. The transformations include affine operations, contrast adjustments, noise injection, cropping, and sharpening. It is commonly used to increase the diversity and size of training data for computer vision models.