Lightweight Attention-Related Mechanisms and Collaborative Design Strategies for Small Traffic Sign Detection: A Structured Review

Abstract

Small traffic sign detection is an important visual perception task in autonomous driving, advanced driver-assistance systems, and intelligent transportation systems. In road scenes, traffic signs are often distant, small, weakly salient, and degraded by low illumination, weather, occlusion, glare, or motion blur. This review synthesizes representative studies on lightweight attention-related mechanisms for small traffic sign detection. Instead of treating attention as an isolated plug-in, it interprets attention-related design as feature-resource allocation within complete detection systems. The review summarizes definitions and challenges, organizes mechanisms into five functional families, evaluates dataset and metric comparability, and discusses ablation and deployment evidence. The analysis shows that reported gains should be interpreted under source protocols because datasets, metrics, baselines, and hardware settings differ across studies. Future work should emphasize scale-aware metrics, standardized ablations, scenario-stratified benchmarks, and reproducible edge-device evaluation.