v0.7.1

Overview

• Workaround for sam3-related issues
• Keep previoust inference_state

v0.7.0

Overview

Implement SAM3 image predictor and video predictor.

v0.6.1

Overview

• Ignore mask outputs with all False

v0.6.0

Overview

• Support SAM2 video predictor
• Make PIL.Image.MAX_IMAGE_PIXELS adjustable by an environment variable

v0.5.1

Overview

• Support SAM2AutomaticMaskGenerator
• Update GitHub Actions

v0.5.0

Overview

• Support SAM2 models
  • qupath-extension-sam v0.6.0 or above is required

v0.4.1

Overview

• Update dependencies.
  • cuda-toolkit from 11.7 to 11.8.
  • gdown from ^4.7.1 to ^5.1.0.
  • torch from ^1.13.1 to ^2.2.2.
  • torchvision from ^0.14.1 to ^0.17.2.

Related issues

• ksugar/qupath-extension-sam#16 by @halqadasi
• #18 by @ArezooGhodsifard.

v0.4.0

Overview

• Support for registering SAM weights from URL (contribution by @constantinpape)
  • ksugar/qupath-extension-sam#8
  • #11

v0.3.1

Overview

• Resolve issues in SAM Automask on Windows
  • ksugar/qupath-extension-sam#6
  • https://forum.image.sc/t/qupath-extension-sam-v0-3-is-released-support-point-and-box-prompts-samautomaticmaskgenerator-etc/83304/10

v0.3.0

Overview

• Support points and multi-mask output (@petebankhead)
• Support SamAutomaticMaskGenerator
• Support MobileSAM
• Add opencv-python to dependencies
• Update README
• Add debug config for vscode