Back

ProVision

Github

ProVision is AI Moneyball for Table Tennis. It transforms match film and tournament footage captured by 2D camera into actionable insights backed by tracked ball trajectories, 3D biomechanics, and Moneyball-grade analytics in minutes.

Key Features

Low-latency CNN + SAM 2 pipeline to localize the ball and estimate trajectories

Low-latency CNN + SAM 2 pipeline to localize the ball and estimate trajectories

ViTPose and YOLO to segment players and reconstruct 3D shot trajectories

ViTPose and YOLO to segment players and reconstruct 3D shot trajectories

Reconstruct player kinematics from 2D video to quantify technique efficiency

Reconstruct player kinematics from 2D video to quantify technique efficiency

Model player similarity and generate matchup-specific recommendations from
historical serve/receive, shot choice, and positioning

Model player similarity and generate matchup-specific recommendations from
historical serve/receive, shot choice, and positioning