ORCHESTRA / MOTION STUDIO SIMULATED DATA
A match is a starting point.
Tune a toy motion signature. See how a repeat changes when the archive has gaps, another view is missing, or a routine pattern looks convincing.
✳ This page generates every sample locally. No live footage, camera data, personal data, external model calls or inference.
MOTION CUESEPARATE WINDOWS
SYNTHETIC ARCHIVE / 12 SAMPLE WINDOWS
One cue. A dozen comparisons.
ILLUSTRATIVE FEATURE SPACE
Motion signature map
TOY SCORES · NOT V-JEPA
SEQUENCE CONSISTENCY
MOTION SIMILARITY →
MINIMUM 78%
Each dot is an invented sample window. Brighter dots clear the selected toy similarity cutoff.
01
MONTH 01 / SAMPLE WINDOW
Window 01 · view A
A generated motion fragment, ready for comparison.
91%TOY SCORE
clears cutoff below cutoff no sample / no score selected window
Ready · generated entirely in this browserLOCAL SAMPLE · NO MODEL CALL
01 / COMPAREMotion can find a repeat.
A toy score sorts similar generated fragments. It is not an output from a trained model.
02 / CHECKContext can change the story.
A repeated route may be ordinary. Scene meaning and sequence order need independent review.
03 / LIMITMissing footage stays unknown.
Coverage gaps block a fair comparison. No demo setting can turn missing evidence into a finding.