Vision ML plugin for Houdini (1.0.1)

Changelog [Mar-07-2024]:

• MediaPipe Hands decoder HDA now has a parm for preferred hand orientation mode (first person / mirror video)
• Improved MediaPipe Pose Compare HDA:
  • measurements in local space instead of world space
  • using joint angles instead of positions (more accurate)
  • support for point weights
• Added new draw lines sample scene demoing MediaPipe hands and Android Sensors running inside a solver

* Known issues:
Currently MediaPipe Pose Compare HDA only works with hands. Body and Face will be supported again soon.

Vision ML plugin for Houdini (1.0.0)

Changelog [Mar-04-2024]:

• Added new "Android Sensors" stream type
• Added new "Rotation from device" hda (rotates an object using sensor data from Android devices)
• Fixed transform issues with MediaPipe Body and MediaPipe Face hda decoders
• Exposed more settings for MediaPipe Body:
  • Num bodies to track
  • Min body presence confidence
  • Min landmark trackers confidence
  • 3D Coordinates
• Overall UI polishes
• Overall documentation polishes (HDA's / Streamer tooltips)
• Added sample hip files for MediaPipe Hands -> Kinefx workflows

Vision ML plugin for Houdini (alpha)

* includes vml-streamer binaries.

First release using a QT based ui with:

• Improved overall stability and performance
• Improved app design/stylesheet
• User friendly webcam device names/resolutions
• Less bugs, faster webcam read speeds
• Documented process for adding custom stream types

Important:
This is an alpha release, some issues are still being addressed.

• Video file mode is still unusable at this time. (Webcam is fine!)
• MediaPipe Body trackers behaving weirdly in Houdini

(Tested on Windows 10 and Linux Fedora 36)