Module layers 

Per-frame layer-learning evidence (originally LayerLearner.add/addNative in TS, now frozen verbatim as the parity oracle at tests/support/reference/layers.ts): screen-space motion discontinuities accumulated across the WHOLE recording. finish() — the one-shot region construction that reads these accumulators — stays with the adapter for now; every per-frame pixel loop lives here.

Accumulation order and arithmetic are those of the historical adapter (f64 sums in the same order, JS Math.hypot for motion norms) so the accumulators match bit-for-bit.

Structs

Field

One motion field over cols × rows analysis cells.

FieldMotion

Learner

Constants

MOVING_COLUMN

A field carries real evidence only once some model moved more than analysis jitter and is reasonably well matched; per-frame and long-baseline evidence are both held to this bar. Mean change over a column’s sampled rows, in analysis grey levels, above which the column moved in that frame (Learner::col_moving). Still content changes by less on a decoded recording; moving content by far more.

Functions

informative