Pilot · research stage
Questions we have not answered yet
We seek to investigate whether sufficiently accurate spatial data and expert review can make geometric obstruction analysis useful for officiating education and review. None of the questions below has been answered.
- Q1Can modeled geometry reproduce controlled camera observations?
- Q2How sensitive are results to position and body-shape assumptions?
- Q3Does the interface improve expert reviewers' understanding?
- Q4What information is missing from ordinary video?
- Q5What additional data would be required for credible game analysis?
Current capabilities
- Three synthetic demonstration scenes
- 3D segment-vs-body obstruction checks on five target samples
- Moving one official at a paused instant and comparing results
- Simplified eye-level and overview perspectives from the same geometry
- Manual keyframe editing, JSON import/export, still export
- Local video reference with manual time offset
Not implemented
- Player tracking or pose estimation from video
- Camera calibration or pixel-to-court mapping
- Gaze, attention or peripheral-vision modeling
- Any foul, call or official assessment
- Live-game use of any kind
- Validation against real observations
A pilot would need
- Controlled recordings with surveyed positions
- Expert officiating reviewers to frame useful questions
- Measured body dimensions for participants
- A documented annotation protocol and error estimates
- Agreement on what the tool should never be used for
There is no application form, partner program or waitlist. No organization has endorsed or evaluated this prototype.