1500m track samples

Cole Hocker

Cole is the best current demo case because the clips force the pipeline to follow one runner through a pack and then through a fast finishing segment.

Compact upper body, quick turnover, and a bigger knee-angle sweep when the finish segment opens up.

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.

Knee sweep

124-136 deg

Average left/right knee angle range across the two processed clips.

Hip bounce

5.9-7.1%

Normalized hip vertical oscillation proxy from pose landmarks.

Readable frames

91-100%

Frames where the pipeline kept enough pose signal to use the stride.

Processed clips

8.66s236 of 260 usable frames

Pack tracking segment

A longer 1500m excerpt with camera motion, broadcast graphics, and pack traffic.

Stride rhythm proxy

7.75

Peak spacing in the lower-body motion signal.

Knee lift proxy

32%

Relative vertical knee travel in the normalized pose.

Pose visibility

81%

Mean landmark visibility across the usable clip.

The tracker keeps enough continuity to read the stride even when the pack tightens around him.
Knee-angle sweep sits around 124 deg, with hip bounce just under 6%.
Feet are the weakest read in this clip, so the stride note leans more on knees, hips, and torso.

Source video

Runner isolation

The selected runner mask over the source segment.

Stride skeleton

Pose landmarks rendered frame by frame.

3.98s120 of 120 usable frames

Finish-phase segment

A shorter 1500m sample where the runner is easier to see and the stride opens up.