Augmentor and albumentations
These are competitors offering overlapping image augmentation functionality, though albumentations is more actively maintained (based on its significantly higher GitHub stars) and optimized for speed with modern deep learning frameworks, while Augmentor provides a more beginner-friendly API.
About Augmentor
mdbloice/Augmentor
Image augmentation library in Python for machine learning.
Provides framework-agnostic pipeline-based augmentation with stochastic probability controls and multi-threading support for efficient batch generation. Supports synchronized augmentation of ground truth masks and segmentation labels through dedicated `Pipeline` and `DataPipeline` classes. Includes generators for Keras and PyTorch integration, enabling on-the-fly augmentation during model training without disk I/O.
About albumentations
albumentations-team/albumentations
Fast and flexible image augmentation library. Paper about the library: https://www.mdpi.com/2078-2489/11/2/125
Provides 70+ spatial and pixel-level transforms with a unified API for multiple computer vision tasks (classification, segmentation, object detection, pose estimation) and data types (images, masks, bounding boxes, keypoints). Optimized for speed with native support for PyTorch and TensorFlow pipelines, achieving consistent benchmark performance as the fastest augmentation library. Applies transformations deterministically across all target types simultaneously, ensuring spatial consistency for tasks requiring synchronized augmentation of related annotations.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work