Starstruk Logo MarkStarstruk

v0.2.0

CVHub520/X-AnyLabelingv0.2.0Aug 9, 2023by CVHub520

AI Summary

This release introduces multiple new AI models focused on medical image segmentation (MobileSAM, MedSAM, SAM-Med2D, LVM-Med) along with specialized medical datasets for polyp, skin lesion, and breast cancer segmentation. It also adds lane detection (CLRNet) and pose estimation (DWPose) capabilities, plus new platform-specific versions for CPU/GPU and Linux systems.

Key Highlights

  • Added 4 new SAM (Segment Anything Model) variants for medical imaging: MobileSAM, MedSAM, SAM-Med2D, and LVM-Med
  • Introduced 3 new medical segmentation datasets/models (Kvasir, ISIC, BUID)
  • New platform-specific releases: CPU, GPU, and Linux variants
  • Added CLRNet for lane line detection and DWPose for full-body pose estimation
  • Fixed YOLOv8 issue and GPU inference result exception bugs

New Features

  • X-AnyLabeling-CPU version
  • X-AnyLabeling-GPU version
  • X-AnyLabeling-Linux-CPU version
  • X-AnyLabeling-Linux-GPU version
  • MobileSAM model integration
  • MedSAM model integration
  • SAM-Med2D model integration
  • LVM-Med model integration
  • Kvasir colorectal polyp segmentation
  • ISIC skin lesion segmentation
  • BUID breast ultrasound cancer segmentation
  • CLRNet lane detection
  • DWPose full-body pose estimation

Full Release Notes

## New versions

- X-AnyLabeling-CPU
- X-AnyLabeling-GPU
- X-AnyLabeling-Linux-CPU
- X-AnyLabeling-Linux-GPU

## New features

- [MobileSAM](https://github.com/ChaoningZhang/MobileSAM)
- [MedSAM](https://github.com/bowang-lab/MedSAM)
- [SAM-Med2D](https://github.com/OpenGVLab/SAM-Med2D)
- [LVM-Med](https://github.com/duyhominhnguyen/LVM-Med)
- - [Kvasir-结直肠息肉分割](https://github.com/CVHub520/X-AnyLabeling/tree/main/assets/examples/kvasir)
- - [ISIC-皮肤病镜像分割](https://github.com/CVHub520/X-AnyLabeling/tree/main/assets/examples/isic)
- - [BUID-超声波乳腺癌分割](https://github.com/CVHub520/X-AnyLabeling/tree/main/assets/examples/buid)
- [CLRNet-车道线检测](https://github.com/CVHub520/X-AnyLabeling/tree/main/assets/examples/lane)
- [DWPose-全身人体姿态估计](https://github.com/IDEA-Research/DWPose)

## What's Changed

- Fix yolov8 issue
- Fix gpu inference result exception