Albumentations is a widely adopted open-source image-augmentation ecosystem for computer vision. It helps engineers and researchers create valid training variation while keeping images, masks, bounding boxes, keypoints, volumes, and video frames synchronized.
AlbumentationsX is the actively developed library in the ecosystem. Install it with OpenCV for headless servers and training environments:
pip install "albumentationsx[headless]"The package preserves the familiar albumentations import path:
import albumentations as A
transform = A.Compose(
[
A.RandomCrop(height=256, width=256),
A.HorizontalFlip(p=0.5),
A.RandomBrightnessContrast(p=0.2),
]
)
augmented = transform(image=image, mask=mask)Use the documentation for guides and API details, or open Explore to test transforms in the browser.
| Component | What it provides |
|---|---|
| AlbumentationsX | Actively developed augmentation pipelines for images and synchronized annotations. |
| Website and documentation | Project homepage, guides, API documentation, blog, benchmarks, and transform reference. |
| Explore | Interactive browser playground for trying every transform on user-uploaded images. |
| Albucore | Optimized atomic image-processing functions with workload-aware backend selection. |
| albu-spec | Structured, typed metadata extracted from AlbumentationsX transforms. |
| Benchmark suite | Reproducible image, multichannel, video, and DataLoader comparisons. |
| Examples | Runnable notebooks for classification, segmentation, detection, keypoints, replay, serialization, and integrations. |
| Albumentations | The MIT-licensed legacy library, preserved for reproducibility, continued use, and forks. |
| 160.7M PyPI downloads |
15.3k GitHub stars |
40k+ GitHub-reported public dependents |
2,270 citing research works |
These figures describe public ecosystem adoption, not paid customers or endorsements. Downloads were checked on pepy.tech, stars and public dependents on GitHub, and research works in the deduplicated adoption snapshot on 22 July 2026. Browse the wider public adoption evidence. Albumentations is a NumFOCUS Affiliated Project.
AlbumentationsX offers two licensing options:
- Commercial license: agree on alternative rights for proprietary applications and services, with coverage for your teams, products, and customer deployments. Your agreement with Albumentations, LLC defines the scope, price, and term. Request a quote or email vladimir@albumentations.ai.
- AGPL-3.0-only: available at no charge, including for commercial use, subject to its terms. Commercial or proprietary status alone does not require a purchase. Read the license guide.
The legacy albumentations
package and repository remain available under the MIT License.
- Read the contribution guide and help improve AlbumentationsX.
- Ask usage questions and meet other practitioners on Discord.
- Follow releases and tutorials through the newsletter.
- Support long-term maintenance through GitHub Sponsors.
Sponsorship supports maintenance; it does not include a commercial license or support agreement.
Code · Docs · Explore · Benchmarks · Sponsor
