Knee sweep
120 deg
Average left/right knee angle range in the processed clip.
Gallery
Middle-distance reference
Jakob Ingebrigtsen
Ingebrigtsen's sample is not the cleanest file in the set. It is useful because the page can show where a reference clip is still worth keeping, but should be treated with caution.
Efficient carriage with moderate knee lift, a stable rhythm proxy, and weaker foot visibility than the cleaner clips.
Stride readout
These are clip-derived signals, not a coaching diagnosis. The point is to show what the pipeline can read and where the video still needs review.
Hip bounce
6.1%
Normalized hip vertical oscillation proxy from pose landmarks.
Readable frames
100%
Frames where the pipeline kept enough pose signal to use the stride.
Processed clips
4.42s132 of 132 usable frames
Middle-distance reference segment
A track clip that works as a reference, but not a pristine one.
Stride rhythm proxy
7.04
Peak spacing in the lower-body motion signal.
Knee lift proxy
33%
Relative vertical knee travel in the normalized pose.
Pose visibility
77%
Mean landmark visibility across the usable clip.
- The clip keeps a readable skeleton, but foot visibility is weaker than the main track samples.
- Knee sweep is the lowest in the gallery set, which makes it a useful contrast case.
- The torso range is wider than the cleanest samples, so this page treats it as a tolerance test.
Runner isolation
The selected runner mask over the source segment.
Stride skeleton
Pose landmarks rendered frame by frame.