albumentations 1.4.3
0
An efficient library for image augmentation, providing extensive transformations to support machine
Contents
An efficient library for image augmentation, providing extensive transformations to support machine learning and computer vision tasks.
Stars: 13344, Watchers: 13344, Forks: 1583, Open Issues: 431The albumentations-team/albumentations
repo was created 5 years ago and the last code push was 6 hours ago.
The project is extremely popular with a mindblowing 13344 github stars!
How to Install albumentations
You can install albumentations using pip
pip install albumentations
or add it to a project with poetry
poetry add albumentations
Package Details
- Author
- Vladimir I. Iglovikov, Mikhail Druzhinin, Alex Parinov, Alexander Buslaev, Eugene Khvedchenya
- License
- MIT
- Homepage
- https://albumentations.ai
- PyPi:
- https://pypi.org/project/albumentations/
- GitHub Repo:
- https://github.com/albumentations-team/albumentations
Classifiers
- Scientific/Engineering/Artificial Intelligence
- Scientific/Engineering/Image Processing
- Software Development/Libraries
- Software Development/Libraries/Python Modules
Related Packages
Errors
A list of common albumentations errors.
Code Examples
Here are some albumentations
code examples and snippets.
GitHub Issues
The albumentations package has 431 open issues on GitHub
- is there a function for keypoints and bounding box corresponding removal
- Several frustrations when starting to work with albumentations on image-to-image problems
- Fix safe rotate targets
- Docs: fix inconsistent description for scale_limit in RandomScale and ShiftScaleRotate.
- Add keep ratio flag to Affine transform
- The mask information was incorrectly converted
- Fix bug in RingingOvershoot transform
- Bug in RingingOvershoot augmentation. "RuntimeWarning: invalid value encountered in true_divide"
- Library failed to import with opencv 4.5.5
- Affine doesn't rotate bounding boxes the same as Rotate
- keypoints augmentation for invisible keypoints
- WIP Fix Affine wrong rotation angle
- InvalidArgumentError: All dimensions except 3 must match. Input 1 has shape [8 48 48 512] and doesn't match input 0 with shape [8 40 40 288].
- Unclear warning when using Perspective transform in ReplayMode
- Corrected off-by-one in random crop.