Starstruk Logo MarkStarstruk

v4.0.0

chenglou/pretextv4.0.0Aug 4, 2026by github-actions[bot]

AI Summary

X-AnyLabeling 4.0.0 is a major release built around a PyQt6 interface, expanded annotation workflows, broader AI model support, and a more reliable cross-platform release pipeline.

Key Highlights

  • Add prompt-free Segment Everything workflow powered by SAM2
  • Upgrade to PyQt6 interface
  • Add configurable application fonts and default shortcuts for drawing tools
  • Harden security against unsafe input, path traversal, and decompression bombs
  • Improve performance by reading image dimensions without decoding full pixel data

New Features

  • Prompt-free Segment Everything workflow powered by SAM2
  • Configurable application fonts
  • Default shortcuts for circle, line, point, and line strip drawing tools

Full Release Notes

> PyPI: https://pypi.org/project/x-anylabeling-cvhub/
> Baidu Cloud: https://pan.baidu.com/s/1pgaw02inCvbEgOme9ajDJA?pwd=e528

> [!NOTE]
> Due to compatibility issues across different systems, if the precompiled version doesn’t work properly on your machine, you can try building and running it from source instead. For details, check out the official installation guide and user documentation.

X-AnyLabeling 4.0.0 is a major release built around a PyQt6 interface, expanded annotation workflows, broader AI model support, and a more reliable cross-platform release pipeline. The beta release notes below retain the full development history.

### 🚀 New Features

- Add a prompt-free Segment Everything workflow powered by the SAM2 Automatic Mask Generator, with grid density, minimum region size, cancellation, and polygon or contour output controls.
- Add configurable application fonts and default shortcuts for circle, line, point, and line strip drawing tools.

### 🐛 Bug Fixes

- Align SAM2 automatic mask generation with the official implementation and bound memory usage through chunked grid processing.
- Harden annotation parsing, model downloads, and export paths against unsafe input and path traversal.
- Save annotation files atomically and restore Pillow decompression bomb protection.
- Preserve shape metadata, attributes, defaults, and canvas updates across label editing and upload workflows.
- Stabilize concurrent Chatbot model refreshes, VQA cancellation, video uploads, and video classifier sidecar validation.
- Skip invalid VOC geometries and load YOLOE embedding models independently of the working directory.

### 🛠️ Improvements

- Read image dimensions without decoding full pixel data.
- Refresh the application branding and menu icon system for a more consistent interface.
- Update package metadata, documentation, and in-app links for the official website.
- Add automated website content synchronization from the main repository.

### 🌟 Contributors

A total of 2 developers contributed to this release.

Thank @Crackvignoule, @CVHub520

**Full Changelog**: https://github.com/CVHub520/X-AnyLabeling/compare/v4.0.0-beta.13...v4.0.0