[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-9615":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":23,"defaultBranch":24,"hasWiki":22,"hasPages":23,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":38,"readmeContent":39,"aiSummary":40,"trendingCount":15,"starSnapshotCount":15,"syncStatus":16,"lastSyncTime":41,"discoverSource":42},9615,"albumentations","albumentations-team\u002Falbumentations","albumentations-team","Fast and flexible image augmentation library. Paper about the library: https:\u002F\u002Fwww.mdpi.com\u002F2078-2489\u002F11\u002F2\u002F125","https:\u002F\u002Falbumentations.ai",null,"Python",15309,1709,1,0,2,5,22,6,44.7,"MIT License",true,false,"main",[26,27,28,29,30,31,32,33,34,35,36,37],"augmentation","deep-learning","detection","fast-augmentations","image-augmentation","image-classification","image-processing","image-segmentation","machine-learning","object-detection","python","segmentation","2026-06-12 02:02:10","# Albumentations\n\n[![PyPI version](https:\u002F\u002Fbadge.fury.io\u002Fpy\u002Falbumentations.svg)](https:\u002F\u002Fbadge.fury.io\u002Fpy\u002Falbumentations)\n![CI](https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002Falbumentations\u002Fworkflows\u002FCI\u002Fbadge.svg)\n[![PyPI Downloads](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fdm\u002Falbumentations.svg?label=PyPI%20downloads)](https:\u002F\u002Fpypi.org\u002Fproject\u002Falbumentations\u002F)\n[![Conda Downloads](https:\u002F\u002Fimg.shields.io\u002Fconda\u002Fdn\u002Fconda-forge\u002Falbumentations.svg?label=Conda%20downloads)](https:\u002F\u002Fanaconda.org\u002Fconda-forge\u002Falbumentations)\n\n> 📣 **Stay updated!** [Subscribe to our newsletter](https:\u002F\u002Falbumentations.ai\u002Fsubscribe) for the latest releases, tutorials, and tips directly from the Albumentations team.\n\n[![License: MIT](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-brightgreen.svg)](https:\u002F\u002Fopensource.org\u002Flicenses\u002FMIT)\n[![Gurubase](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGurubase-Ask%20Albumentations%20Guru-006BFF)](https:\u002F\u002Fgurubase.io\u002Fg\u002Falbumentations)\n\n[Docs](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002F) | [Discord](https:\u002F\u002Fdiscord.gg\u002FAKPrrDYNAt) | [Twitter](https:\u002F\u002Ftwitter.com\u002Falbumentations) | [LinkedIn](https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002F100504475\u002F)\n\n## ⚠️ Important Notice: Albumentations is No Longer Maintained\n\n**This repository is no longer actively maintained.** The last update was in June 2025, and no further bug fixes, features, or compatibility updates will be provided.\n\n### 🚀 Introducing AlbumentationsX - The Future of Albumentations\n\nAll development has moved to **[AlbumentationsX](https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002FAlbumentationsX)**, the next-generation successor to Albumentations.\n\n> **Note:** AlbumentationsX uses dual licensing (AGPL-3.0 \u002F Commercial). The AGPL license has strict copyleft requirements - see details below.\n\n### Your Options Moving Forward\n\n#### 1. **Continue Using Albumentations (MIT License)**\n\n- ✅ **Forever free** for all uses including commercial\n- ✅ **No licensing fees or restrictions**\n- ❌ **No bug fixes** - Even critical bugs won't be addressed\n- ❌ **No new features** - Missing out on performance improvements\n- ❌ **No support** - Issues and questions go unanswered\n- ❌ **No compatibility updates** - May break with new Python\u002FPyTorch versions\n\n**Best for:** Projects that work fine with the current version and don't need updates\n\n#### 2. **Upgrade to AlbumentationsX (Dual Licensed)**\n\n- ✅ **Drop-in replacement** - Same API, just `pip install albumentationsx`\n- ✅ **Active development** - Regular updates and new features\n- ✅ **Bug fixes** - Issues are actively addressed\n- ✅ **Performance improvements** - Faster execution\n- ✅ **Community support** - Active Discord and issue tracking\n- ⚠️ **Dual licensed:**\n  - **AGPL-3.0**: Free ONLY for projects licensed under AGPL-3.0 (not compatible with MIT, Apache, BSD, etc.)\n  - **Commercial License**: Required for proprietary use AND permissive open-source projects\n\n**Best for:** Projects that need ongoing support, updates, and new features\n\n> ⚠️ **AGPL License Warning**: The AGPL-3.0 license is NOT compatible with permissive licenses like MIT, Apache 2.0, or BSD. If your project uses any of these licenses, you CANNOT use the AGPL version of AlbumentationsX - you'll need a commercial license.\n\n### Migration is Simple\n\n```bash\n# Uninstall original\npip uninstall albumentations\n\n# Install AlbumentationsX\npip install albumentationsx\n```\n\nThat's it! Your existing code continues to work without any changes:\n\n```python\nimport albumentations as A  # Same import!\n\ntransform = A.Compose([\n    A.RandomCrop(width=256, height=256),\n    A.HorizontalFlip(p=0.5),\n    A.RandomBrightnessContrast(p=0.2),\n])\n```\n\n### Learn More\n\n- 📦 **AlbumentationsX Repository**: \u003Chttps:\u002F\u002Fgithub.com\u002Falbumentations-team\u002FAlbumentationsX>\n- 💰 **Commercial Licensing**: \u003Chttps:\u002F\u002Falbumentations.ai\u002Fpricing>\n- 💬 **Discord Community**: \u003Chttps:\u002F\u002Fdiscord.gg\u002FAKPrrDYNAt>\n\n---\n\n## Original Albumentations README\n\n## GitAds Sponsored\n\n[![Sponsored by GitAds](https:\u002F\u002Fgitads.dev\u002Fv1\u002Fad-serve?source=albumentations-team\u002Falbumentations@github)](https:\u002F\u002Fgitads.dev\u002Fv1\u002Fad-track?source=albumentations-team\u002Falbumentations@github)\n\nAlbumentations is a Python library for image augmentation. Image augmentation is used in deep learning and computer vision tasks to increase the quality of trained models. The purpose of image augmentation is to create new training samples from the existing data.\n\nHere is an example of how you can apply some [pixel-level](#pixel-level-transforms) augmentations from Albumentations to create new images from the original one:\n![parrot](https:\u002F\u002Fhabrastorage.org\u002Fwebt\u002Fbd\u002Fne\u002Frv\u002Fbdnerv5ctkudmsaznhw4crsdfiw.jpeg)\n\n## Why Albumentations\n\n- **Complete Computer Vision Support**: Works with [all major CV tasks](#i-want-to-use-albumentations-for-the-specific-task-such-as-classification-or-segmentation) including classification, segmentation (semantic & instance), object detection, and pose estimation.\n- **Simple, Unified API**: [One consistent interface](#a-simple-example) for all data types - RGB\u002Fgrayscale\u002Fmultispectral images, masks, bounding boxes, and keypoints.\n- **Rich Augmentation Library**: [70+ high-quality augmentations](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002Freference\u002Fsupported-targets-by-transform\u002F) to enhance your training data.\n- **Fast**: Consistently benchmarked as the [fastest augmentation library](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002Fbenchmarks\u002Fimage-benchmarks\u002F), with optimizations for production use.\n- **Deep Learning Integration**: Works with [PyTorch](https:\u002F\u002Fpytorch.org\u002F), [TensorFlow](https:\u002F\u002Fwww.tensorflow.org\u002F), and other frameworks. Part of the [PyTorch ecosystem](https:\u002F\u002Fpytorch.org\u002Fecosystem\u002F).\n- **Created by Experts**: Built by [developers with deep experience in computer vision and machine learning competitions](#authors).\n\n## Table of contents\n\n- [Albumentations](#albumentations)\n  - [Why Albumentations](#why-albumentations)\n  - [Table of contents](#table-of-contents)\n  - [Authors](#authors)\n    - [Current Maintainer](#current-maintainer)\n    - [Emeritus Core Team Members](#emeritus-core-team-members)\n  - [Installation](#installation)\n  - [Documentation](#documentation)\n  - [A simple example](#a-simple-example)\n  - [Getting started](#getting-started)\n    - [I am new to image augmentation](#i-am-new-to-image-augmentation)\n    - [I want to use Albumentations for the specific task such as classification or segmentation](#i-want-to-use-albumentations-for-the-specific-task-such-as-classification-or-segmentation)\n    - [I want to explore augmentations and see Albumentations in action](#i-want-to-explore-augmentations-and-see-albumentations-in-action)\n  - [Who is using Albumentations](#who-is-using-albumentations)\n    - [See also](#see-also)\n  - [List of augmentations](#list-of-augmentations)\n    - [Pixel-level transforms](#pixel-level-transforms)\n    - [Spatial-level transforms](#spatial-level-transforms)\n  - [A few more examples of **augmentations**](#a-few-more-examples-of-augmentations)\n    - [Semantic segmentation on the Inria dataset](#semantic-segmentation-on-the-inria-dataset)\n    - [Medical imaging](#medical-imaging)\n    - [Object detection and semantic segmentation on the Mapillary Vistas dataset](#object-detection-and-semantic-segmentation-on-the-mapillary-vistas-dataset)\n    - [Keypoints augmentation](#keypoints-augmentation)\n  - [Benchmarking results](#benchmark-results)\n    - [System Information](#system-information)\n    - [Benchmark Parameters](#benchmark-parameters)\n    - [Library Versions](#library-versions)\n  - [Performance Comparison](#performance-comparison)\n  - [Contributing](#contributing)\n  - [Community](#community)\n  - [Citing](#citing)\n\n## Authors\n\n### Current Maintainer\n\n[**Vladimir I. Iglovikov**](https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Figlovikov\u002F) | [Kaggle Grandmaster](https:\u002F\u002Fwww.kaggle.com\u002Figlovikov)\n\n### Emeritus Core Team Members\n\n[**Mikhail Druzhinin**](https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Fmikhail-druzhinin-548229100\u002F) | [Kaggle Expert](https:\u002F\u002Fwww.kaggle.com\u002Fdipetm)\n\n[**Alex Parinov**](https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Falex-parinov\u002F) | [Kaggle Master](https:\u002F\u002Fwww.kaggle.com\u002Fcreafz)\n\n[**Alexander Buslaev**](https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Fal-buslaev\u002F) | [Kaggle Master](https:\u002F\u002Fwww.kaggle.com\u002Falbuslaev)\n\n[**Eugene Khvedchenya**](https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Fcvtalks\u002F) | [Kaggle Grandmaster](https:\u002F\u002Fwww.kaggle.com\u002Fbloodaxe)\n\n## Installation\n\nAlbumentations requires Python 3.9 or higher. To install the latest version from PyPI:\n\n```bash\npip install -U albumentations\n```\n\nOther installation options are described in the [documentation](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002F1-introduction\u002Finstallation\u002F).\n\n## Documentation\n\nThe full documentation is available at **[https:\u002F\u002Falbumentations.ai\u002Fdocs\u002F](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002F)**.\n\n## A simple example\n\n```python\nimport albumentations as A\nimport cv2\n\n# Declare an augmentation pipeline\ntransform = A.Compose([\n    A.RandomCrop(width=256, height=256),\n    A.HorizontalFlip(p=0.5),\n    A.RandomBrightnessContrast(p=0.2),\n])\n\n# Read an image with OpenCV and convert it to the RGB colorspace\nimage = cv2.imread(\"image.jpg\")\nimage = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)\n\n# Augment an image\ntransformed = transform(image=image)\ntransformed_image = transformed[\"image\"]\n```\n\n## Getting started\n\n### I am new to image augmentation\n\nPlease start with the [introduction articles](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002F#learning-path) about why image augmentation is important and how it helps to build better models.\n\n### I want to use Albumentations for the specific task such as classification or segmentation\n\nIf you want to use Albumentations for a specific task such as classification, segmentation, or object detection, refer to the [set of articles](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002F1-introduction\u002Fwhat-are-image-augmentations\u002F) that has an in-depth description of this task. We also have a [list of examples](https:\u002F\u002Falbumentations.ai\u002Fdocs\u002Fexamples\u002F) on applying Albumentations for different use cases.\n\n### I want to explore augmentations and see Albumentations in action\n\nCheck the [online demo of the library](https:\u002F\u002Fexplore.albumentations.ai\u002F). With it, you can apply augmentations to different images and see the result. Also, we have a [list of all available augmentations and their targets](#list-of-augmentations).\n\n## Who is using Albumentations\n\n\u003Ca href=\"https:\u002F\u002Fwww.apple.com\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fapple.jpeg\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fresearch.google\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fgoogle.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fopensource.fb.com\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fmeta_research.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fwww.nvidia.com\u002Fen-us\u002Fresearch\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fnvidia_research.jpeg\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fwww.amazon.science\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Famazon_science.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fopensource.microsoft.com\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fmicrosoft.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fengineering.salesforce.com\u002Fopen-source\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fsalesforce_open_source.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fstability.ai\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fstability.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fwww.ibm.com\u002Fopensource\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fibm.jpeg\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fhuggingface.co\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fhugging_face.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fwww.sony.com\u002Fen\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fsony.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fopensource.alibaba.com\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Falibaba.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fopensource.tencent.com\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Ftencent.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fh2o.ai\u002F\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fwww.albumentations.ai\u002Fassets\u002Findustry\u002Fh2o_ai.png\" width=\"100\"\u002F>\u003C\u002Fa>\n\n### See also\n\n- [A list of papers that cite Albumentations](https:\u002F\u002Fscholar.google.com\u002Fcitations?view_op=view_citation&citation_for_view=vkjh9X0AAAAJ:r0BpntZqJG4C).\n- [Open source projects that use Albumentations](https:\u002F\u002Fgithub.com\u002Falbumentations-team\u002Falbumentations\u002Fnetwork\u002Fdependents?dependent_type=PACKAGE).\n\n## List of augmentations\n\n### Pixel-level transforms\n\nPixel-level transforms will change just an input image and will leave any additional targets such as masks, bounding boxes, and keypoints unchanged. For volumetric data (volumes and 3D masks), these transforms are applied independently to each slice along the Z-axis (depth dimension), maintaining consistency across the volume. The list of pixel-level transforms:\n\n- [AdditiveNoise](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FAdditiveNoise)\n- [AdvancedBlur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FAdvancedBlur)\n- [AutoContrast](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FAutoContrast)\n- [Blur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FBlur)\n- [CLAHE](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCLAHE)\n- [ChannelDropout](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FChannelDropout)\n- [ChannelShuffle](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FChannelShuffle)\n- [ChromaticAberration](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FChromaticAberration)\n- [ColorJitter](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FColorJitter)\n- [Defocus](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FDefocus)\n- [Downscale](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FDownscale)\n- [Emboss](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FEmboss)\n- [Equalize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FEqualize)\n- [FDA](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FFDA)\n- [FancyPCA](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FFancyPCA)\n- [FromFloat](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FFromFloat)\n- [GaussNoise](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FGaussNoise)\n- [GaussianBlur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FGaussianBlur)\n- [GlassBlur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FGlassBlur)\n- [HEStain](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FHEStain)\n- [HistogramMatching](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FHistogramMatching)\n- [HueSaturationValue](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FHueSaturationValue)\n- [ISONoise](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FISONoise)\n- [Illumination](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FIllumination)\n- [ImageCompression](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FImageCompression)\n- [InvertImg](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FInvertImg)\n- [MedianBlur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FMedianBlur)\n- [MotionBlur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FMotionBlur)\n- [MultiplicativeNoise](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FMultiplicativeNoise)\n- [Normalize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FNormalize)\n- [PixelDistributionAdaptation](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPixelDistributionAdaptation)\n- [PlanckianJitter](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPlanckianJitter)\n- [PlasmaBrightnessContrast](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPlasmaBrightnessContrast)\n- [PlasmaShadow](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPlasmaShadow)\n- [Posterize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPosterize)\n- [RGBShift](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRGBShift)\n- [RandomBrightnessContrast](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomBrightnessContrast)\n- [RandomFog](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomFog)\n- [RandomGamma](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomGamma)\n- [RandomGravel](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomGravel)\n- [RandomRain](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomRain)\n- [RandomShadow](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomShadow)\n- [RandomSnow](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomSnow)\n- [RandomSunFlare](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomSunFlare)\n- [RandomToneCurve](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomToneCurve)\n- [RingingOvershoot](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRingingOvershoot)\n- [SaltAndPepper](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSaltAndPepper)\n- [Sharpen](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSharpen)\n- [ShotNoise](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FShotNoise)\n- [Solarize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSolarize)\n- [Spatter](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSpatter)\n- [Superpixels](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSuperpixels)\n- [TextImage](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FTextImage)\n- [ToFloat](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FToFloat)\n- [ToGray](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FToGray)\n- [ToRGB](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FToRGB)\n- [ToSepia](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FToSepia)\n- [UnsharpMask](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FUnsharpMask)\n- [ZoomBlur](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FZoomBlur)\n\n### Spatial-level transforms\n\nSpatial-level transforms will simultaneously change both an input image as well as additional targets such as masks, bounding boxes, and keypoints. For volumetric data (volumes and 3D masks), these transforms are applied independently to each slice along the Z-axis (depth dimension), maintaining consistency across the volume. The following table shows which additional targets are supported by each transform:\n\n- Volume: 3D array of shape (D, H, W) or (D, H, W, C) where D is depth, H is height, W is width, and C is number of channels (optional)\n- Mask3D: Binary or multi-class 3D mask of shape (D, H, W) where each slice represents segmentation for the corresponding volume slice\n\n| Transform                                                                                        | Image | Mask | BBoxes | Keypoints | Volume | Mask3D |\n| ------------------------------------------------------------------------------------------------ | :---: | :--: | :----: | :-------: | :----: | :----: |\n| [Affine](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FAffine)                                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [AtLeastOneBBoxRandomCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FAtLeastOneBBoxRandomCrop) | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [BBoxSafeRandomCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FBBoxSafeRandomCrop)             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [CenterCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCenterCrop)                             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [CoarseDropout](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCoarseDropout)                       | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [ConstrainedCoarseDropout](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FConstrainedCoarseDropout) | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Crop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCrop)                                         | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [CropAndPad](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCropAndPad)                             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [CropNonEmptyMaskIfExists](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCropNonEmptyMaskIfExists) | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [D4](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FD4)                                             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [ElasticTransform](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FElasticTransform)                 | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Erasing](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FErasing)                                   | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [FrequencyMasking](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FFrequencyMasking)                 | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [GridDistortion](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FGridDistortion)                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [GridDropout](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FGridDropout)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [GridElasticDeform](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FGridElasticDeform)               | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [HorizontalFlip](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FHorizontalFlip)                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Lambda](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FLambda)                                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [LongestMaxSize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FLongestMaxSize)                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [MaskDropout](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FMaskDropout)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Morphological](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FMorphological)                       | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Mosaic](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FMosaic)                                     | ✓     | ✓    | ✓      | ✓         |        |        |\n| [NoOp](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FNoOp)                                         | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [OpticalDistortion](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FOpticalDistortion)               | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [OverlayElements](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FOverlayElements)                   | ✓     | ✓    |        |           |        |        |\n| [Pad](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPad)                                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [PadIfNeeded](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPadIfNeeded)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Perspective](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPerspective)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [PiecewiseAffine](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPiecewiseAffine)                   | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [PixelDropout](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPixelDropout)                         | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomCrop)                             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomCropFromBorders](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomCropFromBorders)       | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomCropNearBBox](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomCropNearBBox)             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomGridShuffle](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomGridShuffle)               | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomResizedCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomResizedCrop)               | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomRotate90](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomRotate90)                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomScale](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomScale)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomSizedBBoxSafeCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomSizedBBoxSafeCrop)   | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [RandomSizedCrop](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomSizedCrop)                   | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Resize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FResize)                                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Rotate](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRotate)                                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [SafeRotate](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSafeRotate)                             | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [ShiftScaleRotate](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FShiftScaleRotate)                 | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [SmallestMaxSize](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSmallestMaxSize)                   | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [SquareSymmetry](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FSquareSymmetry)                     | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [ThinPlateSpline](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FThinPlateSpline)                   | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [TimeMasking](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FTimeMasking)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [TimeReverse](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FTimeReverse)                           | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [Transpose](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FTranspose)                               | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [VerticalFlip](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FVerticalFlip)                         | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n| [XYMasking](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FXYMasking)                               | ✓     | ✓    | ✓      | ✓         | ✓      | ✓      |\n\n### 3D transforms\n\n3D transforms operate on volumetric data and can modify both the input volume and associated 3D mask.\n\nWhere:\n\n- Volume: 3D array of shape (D, H, W) or (D, H, W, C) where D is depth, H is height, W is width, and C is number of channels (optional)\n- Mask3D: Binary or multi-class 3D mask of shape (D, H, W) where each slice represents segmentation for the corresponding volume slice\n\n| Transform                                                                      | Volume | Mask3D | Keypoints |\n| ------------------------------------------------------------------------------ | :----: | :----: | :-------: |\n| [CenterCrop3D](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCenterCrop3D)       | ✓      | ✓      | ✓         |\n| [CoarseDropout3D](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCoarseDropout3D) | ✓      | ✓      | ✓         |\n| [CubicSymmetry](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FCubicSymmetry)     | ✓      | ✓      | ✓         |\n| [Pad3D](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPad3D)                     | ✓      | ✓      | ✓         |\n| [PadIfNeeded3D](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FPadIfNeeded3D)     | ✓      | ✓      | ✓         |\n| [RandomCrop3D](https:\u002F\u002Fexplore.albumentations.ai\u002Ftransform\u002FRandomCrop3D)       | ✓      | ✓      | ✓         |\n\n## A few more examples of **augmentations**\n\n### Semantic segmentation on the Inria dataset\n\n![inria](https:\u002F\u002Fhabrastorage.org\u002Fwebt\u002Fsu\u002Fwa\u002Fnp\u002Fsuwanpeo6ww7wpwtobtrzd_cg20.jpeg)\n\n### Medical imaging\n\n![medical](https:\u002F\u002Fhabrastorage.org\u002Fwebt\u002F1i\u002Ffi\u002Fwz\u002F1ifiwzy0lxetc4nwjvss-71nkw0.jpeg)\n\n### Object detection and semantic segmentation on the Mapillary Vistas dataset\n\n![vistas](https:\u002F\u002Fhabrastorage.org\u002Fwebt\u002Frz\u002F-h\u002F3j\u002Frz-h3jalbxic8o_fhucxysts4tc.jpeg)\n\n### Keypoints augmentation\n\n\u003Cimg src=\"https:\u002F\u002Fhabrastorage.org\u002Fwebt\u002Fe-\u002F6k\u002Fz-\u002Fe-6kz-fugp2heak3jzns3bc-r8o.jpeg\" width=100%>\n\n## Benchmark Results\n\n### Image Benchmark Results\n\n### System Information\n\n- Platform: macOS-15.1-arm64-arm-64bit\n- Processor: arm\n- CPU Count: 16\n- Python Version: 3.12.8\n\n### Benchmark Parameters\n\n- Number of images: 2000\n- Runs per transform: 5\n- Max warmup iterations: 1000\n\n### Library Versions\n\n- albumentations: 2.0.4\n- augly: 1.0.0\n- imgaug: 0.4.0\n- kornia: 0.8.0\n- torchvision: 0.20.1\n\n## Performance Comparison\n\nNumber shows how many uint8 images per second can be processed on one CPU thread. Larger is better.\nThe Speedup column shows how many times faster Albumentations is compared to the fastest other\nlibrary for each transform.\n\n| Transform            | albumentations\u003Cbr>2.0.4   | augly\u003Cbr>1.0.0   | imgaug\u003Cbr>0.4.0   | kornia\u003Cbr>0.8.0   | torchvision\u003Cbr>0.20.1   | Speedup\u003Cbr>(Alb\u002Ffastest other)   |\n|:---------------------|:--------------------------|:-----------------|:------------------|:------------------|:------------------------|:---------------------------------|\n| Affine               | **1445 ± 9**              | -                | 1328 ± 16         | 248 ± 6           | 188 ± 2                 | 1.09x                            |\n| AutoContrast         | **1657 ± 13**             | -                | -                 | 541 ± 8           | 344 ± 1                 | 3.06x                            |\n| Blur                 | **7657 ± 114**            | 386 ± 4          | 5381 ± 125        | 265 ± 11          | -                       | 1.42x                            |\n| Brightness           | **11985 ± 455**           | 2108 ± 32        | 1076 ± 32         | 1127 ± 27         | 854 ± 13                | 5.68x                            |\n| CLAHE                | **647 ± 4**               | -                | 555 ± 14          | 165 ± 3           | -                       | 1.17x                            |\n| CenterCrop128        | **119293 ± 2164**         | -                | -                 | -                 | -                       | N\u002FA                              |\n| ChannelDropout       | **11534 ± 306**           | -                | -                 | 2283 ± 24         | -                       | 5.05x                            |\n| ChannelShuffle       | **6772 ± 109**            | -                | 1252 ± 26         | 1328 ± 44         | 4417 ± 234              | 1.53x                            |\n| CoarseDropout        | **18962 ± 1346**          | -                | 1190 ± 22         | -                 | -                       | 15.93x                           |\n| ColorJitter          | **1020 ± 91**             | 418 ± 5          | -                 | 104 ± 4           | 87 ± 1                  | 2.44x                            |\n| Contrast             | **12394 ± 363**           | 1379 ± 25        | 717 ± 5           | 1109 ± 41         | 602 ± 13                | 8.99x                            |\n| CornerIllumination   | **484 ± 7**               | -                | -                 | 452 ± 3           | -                       | 1.07x                            |\n| Elastic              | 374 ± 2                   | -                | **395 ± 14**      | 1 ± 0             | 3 ± 0                   | 0.95x                            |\n| Equalize             | **1236 ± 21**             | -                | 814 ± 11          | 306 ± 1           | 795 ± 3                 | 1.52x                            |\n| Erasing              | **27451 ± 2794**          | -                | -                 | 1210 ± 27         | 3577 ± 49               | 7.67x                            |\n| GaussianBlur         | **2350 ± 118**            | 387 ± 4          | 1460 ± 23         | 254 ± 5           | 127 ± 4                 | 1.61x                            |\n| GaussianIllumination | **720 ± 7**               | -                | -                 | 436 ± 13          | -                       | 1.65x                            |\n| GaussianNoise        | **315 ± 4**               | -                | 263 ± 9           | 125 ± 1           | -                       | 1.20x                            |\n| Grayscale            | **32284 ± 1130**          | 6088 ± 107       | 3100 ± 24         | 1201 ± 52         | 2600 ± 23               | 5.30x                            |\n| HSV                  | **1197 ± 23**             | -                | -                 | -                 | -                       | N\u002FA                              |\n| HorizontalFlip       | **14460 ± 368**           | 8808 ± 1012      | 9599 ± 495        | 1297 ± 13         | 2486 ± 107              | 1.51x                            |\n| Hue                  | **1944 ± 64**             | -                | -                 | 150 ± 1           | -                       | 12.98x                           |\n| Invert               | **27665 ± 3803**          | -                | 3682 ± 79         | 2881 ± 43         | 4244 ± 30               | 6.52x                            |\n| JpegCompression      | **1321 ± 33**             | 1202 ± 19        | 687 ± 26          | 120 ± 1           | 889 ± 7                 | 1.10x                            |\n| LinearIllumination   | 479 ± 5                   | -                | -                 | **708 ± 6**       | -                       | 0.68x                            |\n| MedianBlur           | **1229 ± 9**              | -                | 1152 ± 14         | 6 ± 0             | -                       | 1.07x                            |\n| MotionBlur           | **3521 ± 25**             | -                | 928 ± 37          | 159 ± 1           | -                       | 3.79x                            |\n| Normalize            | **1819 ± 49**             | -                | -                 | 1251 ± 14         | 1018 ± 7                | 1.45x                            |\n| OpticalDistortion    | **661 ± 7**               | -                | -                 | 174 ± 0           | -                       | 3.80x                            |\n| Pad                  | **48589 ± 2059**          | -                | -                 | -                 | 4889 ± 183              | 9.94x                            |\n| Perspective          | **1206 ± 3**              | -                | 908 ± 8           | 154 ± 3           | 147 ± 5                 | 1.33x                            |\n| PlankianJitter       | **3221 ± 63**             | -                | -                 | 2150 ± 52         | -                       | 1.50x                            |\n| PlasmaBrightness     | **168 ± 2**               | -                | -                 | 85 ± 1            | -                       | 1.98x                            |\n| PlasmaContrast       | **145 ± 3**               | -                | -                 | 84 ± 0            | -                       | 1.71x                            |\n| PlasmaShadow         | 183 ± 5                   | -                | -                 | **216 ± 5**       | -                       | 0.85x                            |\n| Posterize            | **12979 ± 1121**          | -                | 3111 ± 95         | 836 ± 30          | 4247 ± 26               | 3.06x                            |\n| RGBShift             | **3391 ± 104**            | -                | -                 | 896 ± 9           | -                       | 3.79x                            |\n| Rain                 | **2043 ± 115**            | -                | -                 | 1493 ± 9          | -                       | 1.37x                            |\n| RandomCrop128        | **111859 ± 1374**         | 45395 ± 934      | 21408 ± 622       | 2946 ± 42         | 31450 ± 249             | 2.46x                            |\n| RandomGamma          | **12444 ± 753**           | -                | 3504 ± 72         | 230 ± 3           | -                       | 3.55x                            |\n| RandomResizedCrop    | **4347 ± 37**             | -                | -                 | 661 ± 16          | 837 ± 37                | 5.19x                            |\n| Resize               | **3532 ± 67**             | 1083 ± 21        | 2995 ± 70         | 645 ± 13          | 260 ± 9                 | 1.18x                            |\n| Rotate               | **2912 ± 68**             | 1739 ± 105       | 2574 ± 10         | 256 ± 2           | 258 ± 4                 | 1.13x                            |\n| SaltAndPepper        | **629 ± 6**               | -                | -                 | 480 ± 12          | -                       | 1.31x                            |\n| Saturation           | **1596 ± 24**             | -                | 495 ± 3           | 155 ± 2           | -                       | 3.22x                            |\n| Sharpen              | **2346 ± 10**             | -                | 1101 ± 30         | 201 ± 2           | 220 ± 3                 | 2.13x                            |\n| Shear                | **1299 ± 11**             | -                | 1244 ± 14         | 261 ± 1           | -                       | 1.04x                            |\n| Snow                 | **611 ± 9**               | -                | -                 | 143 ± 1           | -                       | 4.28x                            |\n| Solarize             | **11756 ± 481**           | -                | 3843 ± 80         | 263 ± 6           | 1032 ± 14               | 3.06x                            |\n| ThinPlateSpline      | **82 ± 1**                | -                | -                 | 58 ± 0            | -                       | 1.41x                            |\n| VerticalFlip         | **32386 ± 936**           | 16830 ± 1653     | 19935 ± 1708      | 2872 ± 37         | 4696 ± 161              | 1.62x                            |\n\n## Contributing\n\nTo create a pull request to the repository, follow the documentation at [CONTRIBUTING.md](CONTRIBUTING.md)\n\n![https:\u002F\u002Fgithub.com\u002Falbuemntations-team\u002Falbumentation\u002Fgraphs\u002Fcontributors](https:\u002F\u002Fcontrib.rocks\u002Fimage?repo=albumentations-team\u002Falbumentations)\n\n## Community\n\n- [LinkedIn](https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002Falbumentations\u002F)\n- [Twitter](https:\u002F\u002Ftwitter.com\u002Falbumentations)\n- [Discord](https:\u002F\u002Fdiscord.gg\u002FAKPrrDYNAt)\n\n## Citing\n\nIf you find this library useful for your research, please consider citing [Albumentations: Fast and Flexible Image Augmentations](https:\u002F\u002Fwww.mdpi.com\u002F2078-2489\u002F11\u002F2\u002F125):\n\n```bibtex\n@Article{info11020125,\n    AUTHOR = {Buslaev, Alexander and Iglovikov, Vladimir I. and Khvedchenya, Eugene and Parinov, Alex and Druzhinin, Mikhail and Kalinin, Alexandr A.},\n    TITLE = {Albumentations: Fast and Flexible Image Augmentations},\n    JOURNAL = {Information},\n    VOLUME = {11},\n    YEAR = {2020},\n    NUMBER = {2},\n    ARTICLE-NUMBER = {125},\n    URL = {https:\u002F\u002Fwww.mdpi.com\u002F2078-2489\u002F11\u002F2\u002F125},\n    ISSN = {2078-2489},\n    DOI = {10.3390\u002Finfo11020125}\n}\n```\n\n---\n\n## 📫 Stay Connected\n\nNever miss updates, tutorials, and tips from the Albumentations team! [Subscribe to our newsletter](https:\u002F\u002Falbumentations.ai\u002Fsubscribe).\n\n\u003C!-- GitAds-Verify: ERJY18YZIFA1WZ6CUJNP379R3T3JN418 -->\n","Albumentations 是一个快速且灵活的图像增强库，专为深度学习任务设计。它支持多种图像处理操作，如分类、检测和分割，并提供高效的执行速度。该库的核心功能包括丰富的图像变换选项、易于使用的API以及与主流深度学习框架的良好兼容性。适用于需要高质量图像数据增强的各种计算机视觉项目，特别是在训练神经网络时提高模型泛化能力的场景中。尽管当前版本已不再维护，但其稳定性和功能性仍能满足许多现有应用需求。对于寻求持续更新和支持的用户，则建议迁移到其后继项目 AlbumentationsX。","2026-06-11 03:23:46","top_topic"]