[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-9677":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":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":47,"readmeContent":48,"aiSummary":49,"trendingCount":16,"starSnapshotCount":16,"syncStatus":50,"lastSyncTime":51,"discoverSource":52},9677,"cvat","cvat-ai\u002Fcvat","cvat-ai","Computer Vision Annotation Tool (CVAT) is a leading platform for building high-quality visual datasets for vision AI. It offers open-source, cloud, and enterprise products, as well as labeling services, for image, video, and 3D annotation with AI-assisted labeling, quality assurance, team collaboration, analytics, and developer APIs.","https:\u002F\u002Fcvat.ai",null,"Python",16037,3702,180,527,0,15,60,226,61,120,"MIT License",false,"develop",true,[27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46],"annotation","annotation-tool","annotations","boundingbox","computer-vision","computer-vision-annotation","dataset","deep-learning","image-annotation","image-classification","image-labeling","image-labelling-tool","imagenet","labeling","labeling-tool","object-detection","pytorch","semantic-segmentation","tensorflow","video-annotation","2026-06-12 04:00:46","\u003Cp align=\"center\">\n  \u003Cimg src=\"\u002Fsite\u002Fcontent\u002Fen\u002Fimages\u002Fcvat-readme-gif.gif\" alt=\"CVAT Platform\" width=\"100%\" max-width=\"800px\">\n\u003C\u002Fp>\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fapp.cvat.ai\u002F\">\n    \u003Cimg src=\"\u002Fsite\u002Fcontent\u002Fen\u002Fimages\u002Fcvat-readme-button-tr-bg.png\" alt=\"Start Annotating Now\">\n  \u003C\u002Fa>\n\u003C\u002Fp>\n\n# Computer Vision Annotation Tool (CVAT)\n\n[![CI][ci-img]][ci-url]\n[![Gitter chat][gitter-img]][gitter-url]\n[![Discord][discord-img]][discord-url]\n[![Coverage Status][coverage-img]][coverage-url]\n[![server pulls][docker-server-pulls-img]][docker-server-image-url]\n[![ui pulls][docker-ui-pulls-img]][docker-ui-image-url]\n[![DOI][doi-img]][doi-url]\n[![Status][status-img]][status-url]\n\nCVAT is an interactive video and image annotation\ntool for computer vision. It is used by tens of thousands of users and\ncompanies around the world. Our mission is to help developers, companies, and\norganizations around the world to solve real problems using the Data-centric\nAI approach.\n\nStart using CVAT online: [cvat.ai](https:\u002F\u002Fcvat.ai). You can use it for free,\nor [subscribe](https:\u002F\u002Fwww.cvat.ai\u002Fpricing\u002Fcloud) to get unlimited data,\norganizations, autoannotations, and [Roboflow and HuggingFace integration](https:\u002F\u002Fwww.cvat.ai\u002Fpost\u002Fintegrating-hugging-face-and-roboflow-models).\n\nOr set CVAT up as a self-hosted solution:\n[Self-hosted Installation Guide](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fadministration\u002Fbasics\u002Finstallation\u002F).\nWe provide [Enterprise support](https:\u002F\u002Fwww.cvat.ai\u002Fpricing\u002Fon-prem) for\nself-hosted installations with premium features: SSO, LDAP, Roboflow and\nHuggingFace integrations, and advanced analytics (coming soon). We also\ndo trainings and a dedicated support with 24 hour SLA.\n\n## Quick start ⚡\n\n- [Installation guide](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fadministration\u002Fbasics\u002Finstallation\u002F)\n- [Manual](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fmanual\u002F)\n- [Contributing](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fcontributing\u002F)\n- [Datumaro dataset framework](https:\u002F\u002Fgithub.com\u002Fcvat-ai\u002Fdatumaro\u002Fblob\u002Fdevelop\u002FREADME.md)\n- [Server API](#api)\n- [Python SDK](#sdk)\n- [Command line tool](#cli)\n- [XML annotation format](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fmanual\u002Fadvanced\u002Fxml_format\u002F)\n- [AWS Deployment Guide](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fadministration\u002Fbasics\u002Faws-deployment-guide\u002F)\n- [Frequently asked questions](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Ffaq\u002F)\n- [Where to ask questions](#where-to-ask-questions)\n\n## Partners ❤️\n\nCVAT is used by teams all over the world. In the list, you can find key companies which\nhelp us support the product or an essential part of our ecosystem. If you use us,\nplease drop us a line at [contact@cvat.ai](mailto:contact+github@cvat.ai).\n\n- [Human Protocol](https:\u002F\u002Fhmt.ai) uses CVAT as a way of adding annotation service to the Human Protocol.\n- [FiftyOne](https:\u002F\u002Ffiftyone.ai) is an open-source dataset curation and model analysis\n  tool for visualizing, exploring, and improving computer vision datasets and models that are\n  [tightly integrated](https:\u002F\u002Fvoxel51.com\u002Fdocs\u002Ffiftyone\u002Fintegrations\u002Fcvat.html) with CVAT\n  for annotation and label refinement.\n\n## Public datasets\n\n[ATLANTIS](https:\u002F\u002Fgithub.com\u002Fsmhassanerfani\u002Fatlantis), an open-source dataset for semantic segmentation\nof waterbody images, developed by [iWERS](http:\u002F\u002Fce.sc.edu\u002Fiwers\u002F) group in the\nDepartment of Civil and Environmental Engineering at the University of South Carolina is using CVAT.\n\nFor developing a semantic segmentation dataset using CVAT, see:\n\n- [ATLANTIS published article](https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fpii\u002FS1364815222000391)\n- [ATLANTIS Development Kit](https:\u002F\u002Fgithub.com\u002Fsmhassanerfani\u002Fatlantis\u002Ftree\u002Fmaster\u002Fadk)\n- [ATLANTIS annotation tutorial videos](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PLIfLGY-zZChS5trt7Lc3MfNhab7OWl2BR).\n\n## CVAT online: [cvat.ai](https:\u002F\u002Fcvat.ai)\n\nThis is an online version of CVAT. It's free, efficient, and easy to use.\n\n[cvat.ai](https:\u002F\u002Fcvat.ai) runs the latest version of the tool. You can create up\nto 10 tasks there and upload up to 500Mb of data to annotate. It will only be\nvisible to you or the people you assign to it.\n\nFor now, it does not have [analytics features](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fadministration\u002Fadvanced\u002Fanalytics\u002F)\nlike management and monitoring the data annotation team. It also does not allow exporting images, just the annotations.\n\nWe plan to enhance [cvat.ai](https:\u002F\u002Fcvat.ai) with new powerful features. Stay tuned!\n\n## Prebuilt Docker images 🐳\n\nPrebuilt docker images are the easiest way to start using CVAT locally. They are available on Docker Hub:\n\n- [cvat\u002Fserver](https:\u002F\u002Fhub.docker.com\u002Fr\u002Fcvat\u002Fserver)\n- [cvat\u002Fui](https:\u002F\u002Fhub.docker.com\u002Fr\u002Fcvat\u002Fui)\n\nThe images have been downloaded more than 1M times so far.\n\n## Screencasts 🎦\n\nHere are some screencasts showing how to use CVAT.\n\n\u003C!--lint disable maximum-line-length-->\n\n[Computer Vision Annotation Course](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PL0to7Ng4PuuYQT4eXlHb_oIlq_RPeuasN):\nwe introduce our course series designed to help you annotate data faster and better\nusing CVAT. This course is about CVAT deployment and integrations, it includes\npresentations and covers the following topics:\n\n- **Speeding up your data annotation process: introduction to CVAT and Datumaro**.\n  What problems do CVAT and Datumaro solve, and how they can speed up your model\n  training process. Some resources you can use to learn more about how to use them.\n- **Deployment and use CVAT**. Use the app online at [app.cvat.ai](https:\u002F\u002Fapp.cvat.ai).\n  A local deployment. A containerized local deployment with Docker Compose (for regular use),\n  and a local cluster deployment with Kubernetes (for enterprise users). A 2-minute\n  tour of the interface, a breakdown of CVAT’s internals, and a demonstration of how\n  to deploy CVAT using Docker Compose.\n\n[Product tour](https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PL0to7Ng4Puua37NJVMIShl_pzqJTigFzg): in this course, we show how to use CVAT, and help to get familiar with CVAT functionality and interfaces. This course does not cover integrations and is dedicated solely to CVAT. It covers the following topics:\n\n- **Pipeline**. In this video, we show how to use [app.cvat.ai](https:\u002F\u002Fapp.cvat.ai): how to sign up, upload your data, annotate it, and download it.\n\n\u003C!--lint enable maximum-line-length-->\n\nFor feedback, please see [Contact us](#contact-us)\n\n## API\n\n- [Documentation](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fapi_sdk\u002Fapi\u002F)\n\n## SDK\n\n- Install with `pip install cvat-sdk`\n- [PyPI package homepage](https:\u002F\u002Fpypi.org\u002Fproject\u002Fcvat-sdk\u002F)\n- [Documentation](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fapi_sdk\u002Fsdk\u002F)\n\n## CLI\n\n- Install with `pip install cvat-cli`\n- [PyPI package homepage](https:\u002F\u002Fpypi.org\u002Fproject\u002Fcvat-cli\u002F)\n- [Documentation](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fapi_sdk\u002Fcli\u002F)\n\n## Supported annotation formats\n\nCVAT supports multiple annotation formats. You can select the format\nafter clicking the **Upload annotation** and **Dump annotation** buttons.\n[Datumaro](https:\u002F\u002Fgithub.com\u002Fcvat-ai\u002Fdatumaro) dataset framework allows\nadditional dataset transformations with its command line tool and Python library.\n\nFor more information about the supported formats, see:\n[Annotation Formats](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fmanual\u002Fadvanced\u002Fformats\u002F).\n\n\u003C!--lint disable maximum-line-length-->\n\n| Annotation format                                                                                | Import | Export |\n| ------------------------------------------------------------------------------------------------ | ------ | ------ |\n| [CVAT for images](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fmanual\u002Fadvanced\u002Fxml_format\u002F#annotation)              | ✔️     | ✔️     |\n| [CVAT for a video](https:\u002F\u002Fdocs.cvat.ai\u002Fdocs\u002Fmanual\u002Fadvanced\u002Fxml_format\u002F#interpolation)          | ✔️     | ✔️     |\n| [Datumaro](https:\u002F\u002Fgithub.com\u002Fcvat-ai\u002Fdatumaro)                                                  | ✔️     | ✔️     |\n| [PASCAL VOC](http:\u002F\u002Fhost.robots.ox.ac.uk\u002Fpascal\u002FVOC\u002F)                                            | ✔️     | ✔️     |\n| Segmentation masks from [PASCAL VOC](http:\u002F\u002Fhost.robots.ox.ac.uk\u002Fpascal\u002FVOC\u002F)                    | ✔️     | ✔️     |\n| [YOLO](https:\u002F\u002Fpjreddie.com\u002Fdarknet\u002Fyolo\u002F)                                                       | ✔️     | ✔️     |\n| [MS COCO Object Detection](http:\u002F\u002Fcocodataset.org\u002F#format-data)                                  | ✔️     | ✔️     |\n| [MS COCO Keypoints Detection](http:\u002F\u002Fcocodataset.org\u002F#format-data)                               | ✔️     | ✔️     |\n| [MOT](https:\u002F\u002Fmotchallenge.net\u002F)                                                                 | ✔️     | ✔️     |\n| [MOTS PNG](https:\u002F\u002Fwww.vision.rwth-aachen.de\u002Fpage\u002Fmots)                                          | ✔️     | ✔️     |\n| [LabelMe 3.0](http:\u002F\u002Flabelme.csail.mit.edu\u002FRelease3.0)                                           | ✔️     | ✔️     |\n| [ImageNet](http:\u002F\u002Fwww.image-net.org)                                                             | ✔️     | ✔️     |\n| [CamVid](http:\u002F\u002Fmi.eng.cam.ac.uk\u002Fresearch\u002Fprojects\u002FVideoRec\u002FCamVid\u002F)                             | ✔️     | ✔️     |\n| [WIDER Face](http:\u002F\u002Fshuoyang1213.me\u002FWIDERFACE\u002F)                                                  | ✔️     | ✔️     |\n| [VGGFace2](https:\u002F\u002Fgithub.com\u002Fox-vgg\u002Fvgg_face2)                                                  | ✔️     | ✔️     |\n| [Market-1501](https:\u002F\u002Fwww.aitribune.com\u002Fdataset\u002F2018051063)                                      | ✔️     | ✔️     |\n| [ICDAR13\u002F15](https:\u002F\u002Frrc.cvc.uab.es\u002F?ch=2)                                                       | ✔️     | ✔️     |\n| [Open Images V6](https:\u002F\u002Fstorage.googleapis.com\u002Fopenimages\u002Fweb\u002Findex.html)                       | ✔️     | ✔️     |\n| [Cityscapes](https:\u002F\u002Fwww.cityscapes-dataset.com\u002Flogin\u002F)                                          | ✔️     | ✔️     |\n| [KITTI](http:\u002F\u002Fwww.cvlibs.net\u002Fdatasets\u002Fkitti\u002F)                                                   | ✔️     | ✔️     |\n| [Kitti Raw Format](https:\u002F\u002Fwww.cvlibs.net\u002Fdatasets\u002Fkitti\u002Fraw_data.php)                           | ✔️     | ✔️     |\n| [LFW](http:\u002F\u002Fvis-www.cs.umass.edu\u002Flfw\u002F)                                                          | ✔️     | ✔️     |\n| [Supervisely Point Cloud Format](https:\u002F\u002Fdocs.supervise.ly\u002Fdata-organization\u002F00_ann_format_navi) | ✔️     | ✔️     |\n| [Ultralytics YOLO Detection](https:\u002F\u002Fdocs.ultralytics.com\u002Fdatasets\u002Fdetect\u002F)                      | ✔️     | ✔️     |\n| [Ultralytics YOLO Oriented Bounding Boxes](https:\u002F\u002Fdocs.ultralytics.com\u002Fdatasets\u002Fobb\u002F)           | ✔️     | ✔️     |\n| [Ultralytics YOLO Segmentation](https:\u002F\u002Fdocs.ultralytics.com\u002Fdatasets\u002Fsegment\u002F)                  | ✔️     | ✔️     |\n| [Ultralytics YOLO Pose](https:\u002F\u002Fdocs.ultralytics.com\u002Fdatasets\u002Fpose\u002F)                             | ✔️     | ✔️     |\n| [Ultralytics YOLO Classification](https:\u002F\u002Fdocs.ultralytics.com\u002Fdatasets\u002Fclassify\u002F)               | ✔️     | ✔️     |\n\n\u003C!--lint enable maximum-line-length-->\n\n## Deep learning serverless functions for automatic labeling\n\nCVAT supports automatic labeling. It can speed up the annotation process\nup to 10x. Here is a list of the algorithms we support, and the platforms they can be run on:\n\n\u003C!--lint disable maximum-line-length-->\n\n| Name                                                                                                    | Type       | Framework  | CPU | GPU |\n| ------------------------------------------------------------------------------------------------------- | ---------- | ---------- | --- | --- |\n| [Segment Anything](\u002Fserverless\u002Fpytorch\u002Ffacebookresearch\u002Fsam\u002Fnuclio\u002F)                                    | interactor | PyTorch    | ✔️  | ✔️  |\n| [Faster RCNN](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fpublic\u002Ffaster_rcnn_inception_resnet_v2_atrous_coco\u002Fnuclio)       | detector   | OpenVINO   | ✔️  |     |\n| [Mask RCNN](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fpublic\u002Fmask_rcnn_inception_resnet_v2_atrous_coco\u002Fnuclio)           | detector   | OpenVINO   | ✔️  |     |\n| [YOLO v3](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fpublic\u002Fyolo-v3-tf\u002Fnuclio)                                            | detector   | OpenVINO   | ✔️  |     |\n| [YOLO v7](\u002Fserverless\u002Fonnx\u002FWongKinYiu\u002Fyolov7\u002Fnuclio)                                                    | detector   | ONNX       | ✔️  | ✔️  |\n| [Object reidentification](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fintel\u002Fperson-reidentification-retail-0277\u002Fnuclio)    | reid       | OpenVINO   | ✔️  |     |\n| [Semantic segmentation for ADAS](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fintel\u002Fsemantic-segmentation-adas-0001\u002Fnuclio) | detector   | OpenVINO   | ✔️  |     |\n| [Text detection v4](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fintel\u002Ftext-detection-0004\u002Fnuclio)                          | detector   | OpenVINO   | ✔️  |     |\n| [SiamMask](\u002Fserverless\u002Fpytorch\u002Ffoolwood\u002Fsiammask\u002Fnuclio)                                                | tracker    | PyTorch    | ✔️  | ✔️  |\n| [TransT](\u002Fserverless\u002Fpytorch\u002Fdschoerk\u002Ftranst\u002Fnuclio)                                                    | tracker    | PyTorch    | ✔️  | ✔️  |\n| [Inside-Outside Guidance](\u002Fserverless\u002Fpytorch\u002Fshiyinzhang\u002Fiog\u002Fnuclio)                                   | interactor | PyTorch    | ✔️  |     |\n| [Faster RCNN](\u002Fserverless\u002Ftensorflow\u002Ffaster_rcnn_inception_v2_coco\u002Fnuclio)                              | detector   | TensorFlow | ✔️  | ✔️  |\n| [RetinaNet](serverless\u002Fpytorch\u002Ffacebookresearch\u002Fdetectron2\u002Fretinanet_r101\u002Fnuclio)                       | detector   | PyTorch    | ✔️  | ✔️  |\n| [Face Detection](\u002Fserverless\u002Fopenvino\u002Fomz\u002Fintel\u002Fface-detection-0205\u002Fnuclio)                             | detector   | OpenVINO   | ✔️  |     |\n\n\u003C!--lint enable maximum-line-length-->\n\n## License\n\nThe code is released under the [MIT License](https:\u002F\u002Fopensource.org\u002Flicenses\u002FMIT).\n\nThe code contained within the `\u002Fserverless` directory is released under the **MIT License**.\nHowever, it may download and utilize various assets, such as source code, architectures, and weights, among others.\nThese assets may be distributed under different licenses, including non-commercial licenses.\nIt is your responsibility to ensure compliance with the terms of these licenses before using the assets.\n\nThis software uses LGPL-licensed libraries from the [FFmpeg](https:\u002F\u002Fwww.ffmpeg.org) project.\nThe exact steps on how FFmpeg was configured and compiled can be found in the [Dockerfile](Dockerfile).\n\nFFmpeg is an open-source framework licensed under LGPL and GPL.\nSee [https:\u002F\u002Fwww.ffmpeg.org\u002Flegal.html](https:\u002F\u002Fwww.ffmpeg.org\u002Flegal.html). You are solely responsible\nfor determining if your use of FFmpeg requires any\nadditional licenses. CVAT.ai Corporation is not responsible for obtaining any\nsuch licenses, nor liable for any licensing fees due in\nconnection with your use of FFmpeg.\n\n## Contact us\n\n[Gitter](https:\u002F\u002Fgitter.im\u002Fopencv-cvat\u002Fpublic) to ask CVAT usage-related questions.\nTypically questions get answered fast by the core team or community. There you can also browse other common questions.\n\n[Discord](https:\u002F\u002Fdiscord.gg\u002FS6sRHhuQ7K) is the place to also ask questions or discuss any other stuff related to CVAT.\n\n[LinkedIn](https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002Fcvat-ai\u002F) for the company and work-related questions.\n\n[YouTube](https:\u002F\u002Fwww.youtube.com\u002F@cvat-ai) to see screencast and tutorials about the CVAT.\n\n[GitHub issues](https:\u002F\u002Fgithub.com\u002Fcvat-ai\u002Fcvat\u002Fissues) for feature requests or bug reports.\nIf it's a bug, please add the steps to reproduce it.\n\n[#cvat](https:\u002F\u002Fstackoverflow.com\u002Fsearch?q=%23cvat) tag on StackOverflow is one more way to ask\nquestions and get our support.\n\n[Use our website](https:\u002F\u002Fwww.cvat.ai\u002Fcontact-us\u002Fenterprise) to reach out to us if you need commercial support.\n\n## Links\n\n- [Intel AI blog: New Computer Vision Tool Accelerates Annotation of Digital Images and Video](https:\u002F\u002Fwww.intel.ai\u002Fintroducing-cvat)\n- [Intel Software: Computer Vision Annotation Tool: A Universal Approach to Data Annotation](https:\u002F\u002Fsoftware.intel.com\u002Fen-us\u002Farticles\u002Fcomputer-vision-annotation-tool-a-universal-approach-to-data-annotation)\n- [VentureBeat: Intel open-sources CVAT, a toolkit for data labeling](https:\u002F\u002Fventurebeat.com\u002F2019\u002F03\u002F05\u002Fintel-open-sources-cvat-a-toolkit-for-data-labeling\u002F)\n- [How to Use CVAT (Roboflow guide)](https:\u002F\u002Fblog.roboflow.com\u002Fcvat\u002F)\n- [How to auto-label data in CVAT with one of 50,000+ models on Roboflow Universe](https:\u002F\u002Fblog.roboflow.com\u002Fhow-to-use-roboflow-models-in-cvat\u002F)\n\n  \u003C!-- Badges -->\n\n[docker-server-pulls-img]: https:\u002F\u002Fimg.shields.io\u002Fdocker\u002Fpulls\u002Fcvat\u002Fserver.svg?style=flat-square&label=server%20pulls\n[docker-server-image-url]: https:\u002F\u002Fhub.docker.com\u002Fr\u002Fcvat\u002Fserver\n[docker-ui-pulls-img]: https:\u002F\u002Fimg.shields.io\u002Fdocker\u002Fpulls\u002Fcvat\u002Fui.svg?style=flat-square&label=UI%20pulls\n[docker-ui-image-url]: https:\u002F\u002Fhub.docker.com\u002Fr\u002Fcvat\u002Fui\n[ci-img]: https:\u002F\u002Fgithub.com\u002Fcvat-ai\u002Fcvat\u002Factions\u002Fworkflows\u002Fmain.yml\u002Fbadge.svg?branch=develop\n[ci-url]: https:\u002F\u002Fgithub.com\u002Fcvat-ai\u002Fcvat\u002Factions\n[gitter-img]: https:\u002F\u002Fimg.shields.io\u002Fgitter\u002Froom\u002Fopencv-cvat\u002Fpublic?style=flat\n[gitter-url]: https:\u002F\u002Fgitter.im\u002Fopencv-cvat\u002Fpublic\n[coverage-img]: https:\u002F\u002Fcodecov.io\u002Fgithub\u002Fcvat-ai\u002Fcvat\u002Fbranch\u002Fdevelop\u002Fgraph\u002Fbadge.svg\n[coverage-url]: https:\u002F\u002Fcodecov.io\u002Fgithub\u002Fcvat-ai\u002Fcvat\n[doi-img]: https:\u002F\u002Fzenodo.org\u002Fbadge\u002F139156354.svg\n[doi-url]: https:\u002F\u002Fzenodo.org\u002Fbadge\u002Flatestdoi\u002F139156354\n[discord-img]: https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F1000789942802337834?label=discord\n[discord-url]: https:\u002F\u002Fdiscord.gg\u002FfNR3eXfk6C\n[status-img]: https:\u002F\u002Fuptime.betterstack.com\u002Fstatus-badges\u002Fv2\u002Fmonitor\u002F1yl3h.svg\n[status-url]: https:\u002F\u002Fstatus.cvat.ai\n","CVAT是一个用于计算机视觉的交互式视频和图像标注工具，广泛应用于机器学习数据标注。其核心功能包括支持多种类型的标注任务（如边界框、语义分割等），并提供了与PyTorch、TensorFlow等深度学习框架的集成能力，以及Roboflow和HuggingFace模型的自动标注功能。该平台适用于任何规模的数据集处理场景，无论是个人开发者还是大型企业团队都能从中受益。此外，CVAT既可以通过云服务直接使用，也支持私有部署，并为后者提供企业级支持选项，包括单点登录、LDAP集成及高级分析等功能。",2,"2026-06-11 03:24:07","top_topic"]