RoadFusion: Latent Diffusion Model for Pavement Defect Detection
This work addresses pavement defect detection for road inspection, offering a novel method to improve robustness in real-world applications.
The paper tackled pavement defect detection by proposing RoadFusion, a framework that uses synthetic anomaly generation and dual-path feature adaptation to address data scarcity and domain shift, achieving state-of-the-art performance on six benchmark datasets.
Pavement defect detection faces critical challenges including limited annotated data, domain shift between training and deployment environments, and high variability in defect appearances across different road conditions. We propose RoadFusion, a framework that addresses these limitations through synthetic anomaly generation with dual-path feature adaptation. A latent diffusion model synthesizes diverse, realistic defects using text prompts and spatial masks, enabling effective training under data scarcity. Two separate feature adaptors specialize representations for normal and anomalous inputs, improving robustness to domain shift and defect variability. A lightweight discriminator learns to distinguish fine-grained defect patterns at the patch level. Evaluated on six benchmark datasets, RoadFusion achieves consistently strong performance across both classification and localization tasks, setting new state-of-the-art in multiple metrics relevant to real-world road inspection.