Stay updated
News & Insights📦 AlbumentationsX 2.0.17
🚀 Major Feature: Oriented Bounding Boxes (OBB) Support
This release introduces native Oriented Bounding Box (OBB) support across a wide range of transforms.
AlbumentationsX now supports both:
- HBB — Horizontal (axis-aligned) bounding boxes
- OBB — Oriented (rotated) bounding boxes
This makes augmentation pipelines consistent for rotated object detection use cases such as aerial imagery, scene text detection, industrial inspection, and autonomous driving.
✔️ Transforms with OBB Support
All links point to the interactive transform explorer:
https://explore.albumentations.ai/transform/<TransformName>
- Affine
- AtLeastOneBBoxRandomCrop
- BBoxSafeRandomCrop
- CenterCrop
- Crop
- CropAndPad
- CropNonEmptyMaskIfExists
- D4
- ElasticTransform
- GridDistortion
- HorizontalFlip
- Lambda
- LongestMaxSize
- Mosaic
- NoOp
- OpticalDistortion
- Pad
- PadIfNeeded
- Perspective
- PiecewiseAffine
- PixelDropout
- RandomCrop
- RandomCropFromBorders
- RandomCropNearBBox
- RandomResizedCrop
- RandomRotate90
- RandomScale
- RandomSizedBBoxSafeCrop
- RandomSizedCrop
- Resize
- Rotate
- SafeRotate
- ShiftScaleRotate
- SmallestMaxSize
- SquareSymmetry
- ThinPlateSpline
- TimeReverse
- Transpose
- VerticalFlip
🧭 New Bounding Box Format
cxcywh (unnormalized)
Added support for: cxcywh
This format:
- Matches YOLO layout
- Uses pixel coordinates
- Is not normalized
Useful for center-based detection workflows.
⚠️ Breaking Changes
In BboxParams:
| Old name | New name |
|---|---|
format | coord_format |
clip | clip_bboxes_on_input |
Please update your pipelines accordingly.
⚡ Performance Improvements (Video)
Speedups when applying transforms to video:
MedianBlurZoomBlurSharpenSolarizeUnsharpMask
These optimizations improve throughput for multi-frame augmentation.
🙌 Contributors
- OBB support &
cxcywh— @ternaus - Video performance — @Dipet
💰 Licensing (Important)
AlbumentationsX is distributed under a dual-license model:
- AGPL — for open-source, academic, and non-commercial use
- Commercial license required — for proprietary, closed-source, or commercial products
If AlbumentationsX is used in a commercial product, including internal training or inference pipelines, you must obtain a commercial license unless you fully comply with AGPL requirements.
👉 Commercial licensing details:
https://albumentations.ai/pricing
❤️ Supporting the Project
AlbumentationsX is widely used, but maintained by a small core team. If AlbumentationsX is useful to you, here are ways you can help:
💙 GitHub Sponsors
Support maintenance, CI, reviews, and features with monthly contributions:
👉 https://github.com/sponsors/albumentations-team
🏢 Commercial License
If AlbumentationsX is part of a proprietary or revenue-generating product, a commercial license:
- Keeps you legally safe
- Directly funds ongoing development (including features like OBB support)
Thank you to everyone who uses, reports issues, sponsors, or supports AlbumentationsX. This release exists because of that support.