Pose Training with Computer Vision

Computer Vision Tools
  • PoseNet
    • pre-trained pose detection via webcam in the browser using mL5.js (a library built on top of TensorFlow.js)
    • used to detect body data via “keypoints” and “skeleton”
    • Coding Train Tutorial
  • Kinect (Azure and v2)
    • externally mounted all-in-one sensor with:
      • rgb webcam for video
      • infrared DepthKit for proximity sensing and volumetric capture
      • skeleton tracking
    • when used with the Kinectron development kit (Windows only), can be used to track body data via “keypoints” and “skeleton”
Pose Training Tools

Comments

Leave a Reply