YOLO-ROC: A High-Precision and Ultra-Lightweight Model for Real-Time Road Damage Detection

2025-07-30

AI Summary

Road damage detection faces two key challenges: insufficient multi-scale feature extraction—leading to missed detections of small objects—and high computational overhead, hindering real-time deployment. To address these, we propose YOLO-ROC, a lightweight real-time detector. It introduces a Bidirectional Multi-Scale Spatial Pyramid Pooling module (BMS-SPPF) integrated with spatial-channel bidirectional attention to enhance small-object feature representation. Additionally, a hierarchical channel compression strategy is adopted to drastically reduce parameter count and computational complexity. Evaluated on the RDD2022_China_Drone dataset, YOLO-ROC achieves an mAP₅₀ of 67.6%, with a 16.8% improvement in detection performance for the small-object class D40. The model contains only 0.89M parameters, occupies 2.0 MB storage, and requires just 2.6 GFLOPs, demonstrating superior accuracy, strong generalization, and efficient edge-device deployability.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Learning on the Edge & Model Compression
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Road damage detection is a critical task for ensuring traffic safety and maintaining infrastructure integrity. While deep learning-based detection methods are now widely adopted, they still face two core challenges: first, the inadequate multi-scale feature extraction capabilities of existing networks for diverse targets like cracks and potholes, leading to high miss rates for small-scale damage; and second, the substantial parameter counts and computational demands of mainstream models, which hinder their deployment for efficient, real-time detection in practical applications. To address these issues, this paper proposes a high-precision and lightweight model, YOLO - Road Orthogonal Compact (YOLO-ROC). We designed a Bidirectional Multi-scale Spatial Pyramid Pooling Fast (BMS-SPPF) module to enhance multi-scale feature extraction and implemented a hierarchical channel compression strategy to reduce computational complexity. The BMS-SPPF module leverages a bidirectional spatial-channel attention mechanism to improve the detection of small targets. Concurrently, the channel compression strategy reduces the parameter count from 3.01M to 0.89M and GFLOPs from 8.1 to 2.6. Experiments on the RDD2022_China_Drone dataset demonstrate that YOLO-ROC achieves a mAP50 of 67.6%, surpassing the baseline YOLOv8n by 2.11%. Notably, the mAP50 for the small-target D40 category improved by 16.8%, and the final model size is only 2.0 MB. Furthermore, the model exhibits excellent generalization performance on the RDD2022_China_Motorbike dataset.

Problem

• Inadequate multi-scale feature extraction for road damage detection
• High computational demands hinder real-time deployment
• Low detection accuracy for small-scale road damage

Innovation

• Bidirectional Multi-scale Spatial Pyramid Pooling Fast module
• Hierarchical channel compression strategy
• Ultra-lightweight model with 2.0 MB size