Project 36 of 39 · AI systems
MediaPipe & OpenCV Experiments
Computer vision exercises with MediaPipe and OpenCV covering hand pose and face mesh tracking.
01Overview
A set of computer vision exercises covering hand pose and face mesh tracking, built while learning how the tracking pipelines behave on real input.
Kept because the notes on what fails, and why, are more useful than the demos themselves.
02How it is built
- Hand pose tracking from a live camera feed.
- Face mesh detection and landmark extraction.
- Frame-by-frame processing with the results drawn back onto the image.