Sports tech sandbox
A simulated study of pose tracking and badminton motion. All values are illustrative.
- Wrist / racket grip
- Right shoulder
- Left shoulder
- Spine core
- Right knee
- Left knee
- Right ankle
- Left ankle
Simulated telemetry
Speed changes with the frame; the other values are fixed examples.
- Wrist / racket speed
- 317.9 km/h
- Model accuracy
- 94.8%
- Latency
- 12.4 ms
- Keypoint target
- 18 target · 8 drawn
- Shot probability
- 88% smash
- Racket angle
- -18.2° steep
30 frames, 150 ms per step. Moving the slider pauses playback. Arrow keys step through frames; Home returns to frame 1.
Static overhead view at frame 1. View and frame controls need JavaScript.
Research directions
A badminton pose-network concept associated with AI & ML study at Woxsen University (2025–2028).
- Multi-person joint positions and pose estimation.
- Racket acceleration tracking and a 300 km/h shuttle-motion target.
- Footwork recovery lag and shoulder rotational velocity.
- Match-shot probability and biomechanical chain analysis.