Pose Estimation

A real-time pose estimation demo running entirely in the browser via TensorFlow.js and PoseNet. The model tracks 17 body keypoints — including eyes, shoulders, elbows, wrists, hips, knees, and ankles — overlaying a skeleton on the camera feed in real time.

No server required; all inference happens client-side. Built as a companion exploration to the Object Detection demo, pushing the boundaries of what machine learning can do directly in a browser.