Work

SCCA-YOLO: A Spatial and Channel Collaborative Attention Enhanced YOLO Network for Highway Autonomous Driving Perception System

Year: 2025

Type: article

Abstract: In the domain of autonomous driving perception systems, where extensive research has been conducted on urban environments, the investigation of rural scenarios has gained prominence due to the overarching objective of achieving comprehensive autonomous driving capabilities. The inherent complexity and unpredictability associated with rural road conditions present distinct challenges for autonomous driving technologies. Consequently, this study introduces a spatial channel collaborative attention YOLO network specifically designed for rural road contexts. This network incorporates an innovative attention mechanism that integrates spatial attention with shared semantics and channel self-attention in a sequential manner, thereby enhancing the accuracy of YOLOv8. Furthermore, the integration of the Ghost module is employed to facilitate the network's lightweight characteristics. Our evaluations, conducted on both proprietary and publicly available datasets, demonstrate the effectiveness of our detection network's performance.

Source: Scientific Reports

DOI: 10.1038/s41598-025-90743-4

Authors: Fengchen Wei, Weiji Wang

Institution: University of Sussex

Language: English

Topic: Advanced Neural Network Applications

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: gold

Estimated APC paid: $2,690