DAPONet: A Dual Attention and Partially Overparameterized Network for Real-Time Road Damage Detection

TL;DR

DAPONet is a novel real-time road damage detection model that combines dual attention, partial over-parameterization, and efficient downsampling to outperform existing methods in accuracy and efficiency.

Contribution

It introduces a dual attention and partial over-parameterization approach specifically designed for real-time road damage detection, achieving significant improvements in speed and accuracy.

Findings

Achieves 70.1% mAP50 on SVRDD, outperforming YOLOv10n by 10.4%.

Reduces parameters by 41% and FLOPs by 80%.

On MS COCO2017, surpasses EfficientDet-D1 with 0.8% higher mAP50-95 and 74% fewer parameters and FLOPs.

Abstract

Current road damage detection methods, relying on manual inspections or sensor-mounted vehicles, are inefficient, limited in coverage, and often inaccurate, especially for minor damages, leading to delays and safety hazards. To address these issues and enhance real-time road damage detection using street view image data (SVRDD), we propose DAPONet, a model incorporating three key modules: a dual attention mechanism combining global and local attention, a multi-scale partial over-parameterization module, and an efficient downsampling module. DAPONet achieves a mAP50 of 70.1% on the SVRDD dataset, outperforming YOLOv10n by 10.4%, while reducing parameters to 1.6M and FLOPs to 1.7G, representing reductions of 41% and 80%, respectively. On the MS COCO2017 val dataset, DAPONet achieves an mAP50-95 of 33.4%, 0.8% higher than EfficientDet-D1, with a 74% reduction in both parameters and FLOPs.