🛣️ YOLOv12s Road Damage Detection (RDD2022)

Demo for rezzzq/yolo12s-road-damage-rdd2022 — a YOLOv12-small model (with A2C2f attention modules) fine-tuned on the RDD2022 dataset. Upload a photo of a road surface to detect damage.

Classes: D00 (longitudinal crack) · D10 (transverse crack) · D20 (alligator crack) · D40 (pothole) · Repair (repaired area)
0.05 0.9

Trained on RDD2022 — see the model card for training details and citation.