docs: record fail-closed accuracy challenger

2026-07-15 05:42:17 +02:00
parent f82a7f456a
commit ecc9f77732
11 changed files with 218 additions and 22 deletions
@@ -183,6 +183,7 @@ Before creating tiles, the guided action inspects raster dimensions and estimate
 - When `yolo-configured` is selected, users can provide an existing raster tile manifest path.
 - Detection Lab lists local model assets from `GET /api/v1/detection/model-assets` so operators can choose an existing mounted model file instead of editing only one hidden `YOLO_MODEL_PATH` slot.
 - Detection Lab exposes explicit operator profiles for mounted local building detectors. The focused small-building model is the recommended recall-balanced `0.15` profile; the previous expanded-AOI `0.15` model remains available for higher precision, and the background-aware `0.35` model remains the conservative review choice. Applying a profile never downloads weights, changes runtime environment or starts inference automatically.
+- The recommended profile metrics are coverage-aligned with the current QA pipeline: mean precision `0.6141`, recall `0.6062` and F1 `0.6069` over seven independent Mol/Kempen zones, with zero detections in all three pure-empty controls. A higher-positive-F1 challenger remains hidden from approved profiles because it failed the Postel-bos empty-control gate.
 - Applying a profile deliberately selects the local model asset and threshold for the browser-run request; runtime default activation remains a separate guarded `.env` operation through `scripts/activate_promoted_yolo_candidate.py`.
 - Detection Lab includes a read-only YOLO runtime preflight panel with backend status, dependency visibility, local model configuration, `torch`/`ultralytics` versions, CUDA state and `YOLO_CONFIG_DIR`.
 - The UI still does not download models or create fake detections; backend status and error codes remain the source of truth.