Accessible feedback
Give casual players a practical way to review one swing without specialized capture equipment.
About KinetiCourt
A tennis project built around one useful next step.
KinetiCourt began as an AP Capstone project asking whether pose estimation could turn an ordinary tennis clip into feedback a casual player can understand and use.
The project
From video frames to a focused cue.
The analysis pipeline uses OpenCV and MediaPipe to extract pose landmarks from tennis video. It summarizes movement into clip-level features, compares those features with a trained reference model, and produces a form score when pose quality is sufficient.
The project also tests racket and ball tracking, but KinetiCourt does not present contact claims unless that tracking passes its reliability checks.
The public beta keeps access narrow while the analysis, privacy controls, and coaching language are tested against real clips.
Project priorities
Honest uncertainty
Withhold scores and contact claims when the input or tracking is not reliable enough.
Private beta
Help test KinetiCourt with real tennis.
Request beta access and add your email to the list.