Download Latest Version
v2.64.0 source code.tar.gz (94.4 MB)
Get an email when there's a new version of Computer Vision Annotation Tool (CVAT)
| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2026-03-17 | 1.4 kB | |
| v2.60.0 source code.tar.gz | 2026-03-17 | 90.9 MB | |
| v2.60.0 source code.zip | 2026-03-17 | 92.2 MB | |
| Totals: 3 Items | 183.1 MB | 1 | |
Added
-
Docker + compose for YOLO (https://github.com/cvat-ai/cvat/pull/10325)
-
Interactors can now return multiple shapes instead of a single one (https://github.com/cvat-ai/cvat/pull/10329)
-
Interactors can now return a confidence attribute. If present, interaction results can be filtered in the UI (https://github.com/cvat-ai/cvat/pull/10329)
-
Added UI controls to remove point and rectangle prompts when using Interactors or OpenCV tools (https://github.com/cvat-ai/cvat/pull/10329)
-
Server may add text prompts to interactors as
text_promptsfield inPOST: /api/lambda/functions/<id>(https://github.com/cvat-ai/cvat/pull/10329)
Changed
-
Unified the expected interactor interface to align with the
/annotationsAPI and AI detector outputs (https://github.com/cvat-ai/cvat/pull/10329) -
Updated the IoG serverless function to support the new interactor interface (https://github.com/cvat-ai/cvat/pull/10329)
Fixed
-
Reduced memory usage by replacing JavaScript arrays with typed arrays (https://github.com/cvat-ai/cvat/pull/10329)
-
Fixed memory leaks related to unreleased object URLs in
cvat-canvas(https://github.com/cvat-ai/cvat/pull/10329) -
[SDK] Fixed a crash in
TasksRepo.create_from_backup,ProjectsRepo.create_from_backup,Task.upload_datathat could occur if a recoverable error occurred during chunk uploading (https://github.com/cvat-ai/cvat/pull/10375)
