rdd2022

YOLOv7-based road damage detection system trained on RDD2022 multi-national dataset. Achieves 63.85% mAP for detecting cracks and potholes. Includes Gradio UI.

PyTorch implementation of EfficientDet (D0, D2, D3) for Road Damage Detection on RDD2022 dataset. Includes K-Medoids anchors, DDP training, and COCO evaluation.

Reproducible CPU-only road damage detection on RDD2022 with domain-aware COCO evaluation and an IEEE paper

RoadVision: Real-time road damage detection with YOLO + Vision Transformers (RDD2022)

AI road damage detection, geolocation & mapping — YOLO26 / RDD-YOLO trained on RDD2022 + 9 public datasets (62.45% mAP@50). FastAPI + React + OpenStreetMap; live in-browser demo with ONNX Runtime Web.

Road damage detection with YOLO26 on RDD2022: cracks and potholes across countries, with a Norway fine-tuning study.

AI road damage detection with severity scoring, cross-frame video tracking and geospatial mapping. YOLO + Streamlit, CPU-only inference.

YOLOv8 pavement distress detection with Grad-CAM explainability — heatmaps, per-class confidence analysis, confusion matrix, and Streamlit dashboard

AI road-damage detection and inspection-priority platform (RDD2022 + YOLO26n)