albumentations

What it is and what it does

Albumentations is a Python library for augmenting training data in computer vision and deep learning tasks. It provides a unified API to apply transformations to images, masks, bounding boxes, and keypoints—supporting classification, semantic and instance segmentation, object detection, and pose estimation. The library separates pixel-level transforms from spatial transforms and applies them consistently across all data types in a single pipeline.

The package depends on numpy, scipy, PyYAML, pydantic, albucore, opencv-python-headless, typing-extensions, and eval-type-backport. It is classified as Production/Stable and requires Python 3.9 or higher. The last release was 444 days ago.

Use it for:

• Augment training datasets for image classification to improve model generalization without collecting new labeled data.
• Apply consistent transformations to images and bounding boxes for object detection model training.
• Generate varied training samples for semantic segmentation by transforming both images and pixel-level masks.
• Augment keypoint annotations alongside images for pose estimation and human pose recognition tasks.
• Preprocess medical imaging data with specialized augmentations for anomaly detection and diagnostic tasks.
• Create synthetic training variations for autonomous driving datasets with coordinated image and bounding-box transforms.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Albumentations applies image transformations to training data, supporting classification, segmentation, object detection, and pose estimation with a unified API for images, masks, bounding boxes, and keypoints.

Yes. Albumentations is a mature, permissively licensed library with low install friction and no known vulnerabilities. The aging maintenance status (444 days since release) is a minor concern but does not block use for stable augmentation pipelines. Install it if you need a unified, production-grade augmentation API for computer vision tasks.

Install

albumentations on PyPI

pip install albumentations

uv add albumentations

poetry add albumentations

Installing albumentations

Before you install

Low friction install with 8 runtime dependencies including numpy, scipy, and opencv-python-headless. Maintenance status is aging (444 days since last release), though the package is marked Production/Stable and supports Python 3.9, 3.10, 3.11, 3.12, 3.13.

License in practice

MIT License permits unrestricted use, modification, and distribution with minimal restrictions—suitable for both open-source and commercial projects.

Quickstart

pip install albumentations

import albumentations as A

transform = A.Compose([
    A.RandomCrop(width=256, height=256),
    A.HorizontalFlip(p=0.5),
])

transformed = transform(image=image)
transformed_image = transformed["image"]

Requires Python 3.9 or higher.

Verify before relying

• Exact number of augmentation transforms available (description mentions '70+' but version-specific count unknown)
• Performance benchmarking details and comparison methodology against other libraries
• Whether all advertised framework integrations are actively maintained in version 2.0.8

Package facts

Evidence: albumentations-2.0.8-py3-none-any.whl

Keywords: 2D augmentation, 3D augmentation, aerial photography, anomaly detection, artificial intelligence, autonomous driving, bounding boxes, classification, computer vision, computer vision library, data augmentation, data preprocessing, data science, deep learning, deep learning library, depth estimation, face recognition, fast augmentation, image augmentation, image processing, image transformation, images, instance segmentation, keras, keypoint detection, keypoints, machine learning, machine learning tools, masks, medical imaging, microscopy, object counting, object detection, optimized performance, panoptic segmentation, pose estimation, python library, pytorch, quality inspection, real-time processing, robotics vision, satellite imagery, semantic segmentation, tensorflow, volumes, volumetric data, volumetric masks

Tags

More Libraries packages