Computer vision / 2026 / Runs natively
Native Fusion Control
Control Windows with your body: Python does face, hand and pose fusion, C++ injects input through SendInput at low latency. No browser anywhere.
01
The problem
Browser-based tracking can't drive the operating system, and Python input injection is too slow to feel real.
02
How I approached it
Split by strength, Python for tracking, fusion, smoothing and gesture logic; a C++ bridge for low-latency SendInput injection, with YOLO11 pose running natively for accuracy the browser can't reach.
03
The outcome
Gesture control that moves the actual cursor without perceptible lag.
Related work
ASL fingerspelling to text from a webcam. It reads hand geometry instead of pixels, so all 26 letters work with no training, and two seconds of calibration per letter fits it to your own hand.
A gesture control layer for hand tracking, built as an Electron app over Python ML backends and now running its tracking front end in a browser too.
A webcam filter that reads your face and hands, picks a reaction pose, and stamps the artwork over your head, then publishes it to a virtual camera Meet, Zoom, Teams and Discord can select.