Upload README.md with huggingface_hub

README.md

@@ -243,11 +243,27 @@ so the numbers below are re-derivable from `benchmark/records/`:
 | painted-coverage MAE | 0.162 |
 
 The all-`other` floor is mIoU **0.110**, so the model clears the floor by a wide
-margin. On the hard substrate/condition questions it is **not** Q4: on 117
-reviewed *wild* photographs the wallpaper head's precision is ~0.50 and the
-drywall head still calls ~50% of painted walls "bare board" — that gap is the
-next stage of work (`docs/ROADMAP.md`, Phase 1) and is stated rather than
-hidden. `train.py` reports the same metric families on any dataset you supply.
+margin. Two readings matter, and both are stated rather than hidden.
+
+**On its own validation data it looks strong:** painted-coverage error is about
+**7.5 points**, and it never marks a wall complete that is not (false-complete
+rate **0.0**).
+
+**On 117 real photographs it was never tuned on, the substrate and condition
+questions are not solved** — the honest half:
+
+| Question (117 wild photos) | Result |
+| --- | --- |
+| Finding the wall | **117 / 117** |
+| Wallpaper precision (22 positives) | **0.50** — half its alarms are false |
+| Painted wall called "bare board" (drywall) | **50%** raw · **65%** calibrated |
+| Damage (peeling / rust), AUROC | **0.52** — close to a coin flip; it flags most sound walls |
+| Sign-off on bare drywall | **1 of 19** walls marked finished — a real robot would leave that wall unpainted |
+
+The quickest gain is fixing those real-photo numbers: **hard negatives for
+painted vs drywall/wallpaper**, or swapping in **SAM 2 / Depth Anything**
+(`docs/ROADMAP.md`, Phase 1–2). `train.py` reports the same metric families on
+any dataset you supply.
 
 ## Intended use and limitations