Welcome to Albumentations documentation¶
Albumentations is a fast and flexible image augmentation library. The library is widely used in industry, deep learning research, machine learning competitions, and open source projects. Albumentations is written in Python, and it is licensed under the MIT license. The source code is available at https://github.com/albumentations-team/albumentations.
If you are new to image augmentation, start with our "Learning Path" for beginners. It describes what image augmentation is, how it can boost deep neural networks' performance, and why you should use Albumentations.
For hands-on experience, check out our "Quick Start Guide" and "Examples" sections. They show how you can use the library for different computer vision tasks: image classification, semantic segmentation, instance segmentation, object detection, and keypoint detection. Each example includes a link to Google Colab, where you can run the code by yourself.
You can also visit explore.albumentations.ai to visually explore and experiment with different augmentations in your browser. This interactive tool helps you better understand how each transform affects images before implementing it in your code.
"API Reference" contains the description of Albumentations' methods and classes.
Quick Start Guide¶
Working with Multi-dimensional Data¶
Volumetric Data (3D)¶
Video and Sequential Data¶
Learning Path¶
Beginners¶
Intermediate¶
Advanced¶
Framework Integration¶
Library Comparisons¶
- Transform Library Comparison - Find equivalent transforms between Albumentations and other libraries (torchvision, Kornia)
- Migration from torchvision - Step-by-step migration guide
Examples¶
- Defining a simple augmentation pipeline for image augmentation
- Using Albumentations to augment bounding boxes for object detection tasks
- How to use Albumentations for detection tasks if you need to keep all bounding boxes
- Using Albumentations for a semantic segmentation task
- Using Albumentations to augment keypoints
- Applying the same augmentation with the same parameters to multiple images, masks, bounding boxes, or keypoints
- Weather augmentations in Albumentations
- Example of applying XYMasking transform
- Example of applying ChromaticAberration transform
- Example of applying Morphological transform
- Example of applying D4 transform
- Example of applying RandomGridShuffle transform
- Example of applying OverlayElements transform
- Example of applying TextImage transform
- Debugging an augmentation pipeline with ReplayCompose
- How to save and load parameters of an augmentation pipeline
- Showcase. Cool augmentation examples on diverse set of images from various real-world tasks.
- How to save and load transforms to HuggingFace Hub.
Examples of how to use Albumentations with different deep learning frameworks¶
- PyTorch and Albumentations for image classification
- PyTorch and Albumentations for semantic segmentation
External resources¶
- Blog posts, podcasts, talks, and videos about Albumentations
- Books that mention Albumentations
- Online courses that cover Albumentations