Model changes are releases.
A safety camera system cannot be judged by a single accuracy number. VESU measures coverage, evidence completeness, publication latency, false-positive burden, replay gate results, and operator burden. Strategies earn promotion through replay and shadow evidence — not because a demo looked better.
No single accuracy number. Six categories, watched continuously.
What fraction of monitored cameras delivered usable observation time, broken down by camera-health states.
Every published incident must carry clip, reason, coverage state, and provenance. Missing fields are a defect.
From Stage 2 candidate to ATMS delivery, by class. Operator-relevant — not just a model-call timer.
Per-class FP per camera-day. Tracks whether the system is teaching operators to mute it.
A candidate strategy must beat the incumbent across replay corpora — by class — before promotion.
When the low-cost stage and the expensive stage disagree often, something is drifting. The disagreement rate is itself a signal.
Five steps. Promotion is earned.
A VESU strategy is a versioned bundle, not a setting. The lifecycle below is the same for every change that affects what operators see.
Roads change. Cameras change. Models change.
VESU watches for that and degrades coverage early — rather than publishing alerts with the wrong reality model.
The current frame's structure is compared to the learned scene. Divergence triggers re-learning.
If observed motion no longer matches the learned direction, the scene model is flagged as stale.
If Stage 3 retracts a high rate of Stage 2 candidates, something has shifted in perception or the world.
Confirmed, rejected, and ambiguous events feed the evaluation set. Patterns become hard negatives.
Observe. Replay. Compare. Promote. Monitor.
Reliability is what makes the rest credible.
We are happy to walk through how a strategy change reaches operators — and what would have to be true for it to be rolled back automatically.