飞控与探测2026，Vol.9Issue(1)：119-130,12.DOI:10.20249/j.cnki.2096-5974.2026.01.010
基于改进YOLO11s的跨模态学习红外目标检测方法
A Cross-Modal Learning Approach to Infrared Object Detection Method Based on Improved YOLO11s
摘要
Abstract
Addressing the core challenge of model severely degraded detection accuracy in infrared object detection due to scarce training data and insufficient annotations,this paper proposes a cross-modal learning-based infrared target detection method based on an improved YOLO11s.The method mitigates the challenge of scarce infrared data by leveraging the abundance of visible-light data.Specifically,it incorporates:A combination of diverse data augmentation strategies to simu-late the infrared domain distribution;A lightweight module that reduces model complexity and op-timizes deployment on resource-constrained edge devices;Integration of a partial self-attention mechanism after the C2 PSA module to optimize visible-light feature representation,thereby im-proving the mode'l s cross-modal generalization capability for the infrared domain.Experimental results demonstrate that the proposed method achieves the mAP of 52.2％,while maintaining a parame-ter count of 8.8×106 on the Drone Vehicle dataset.Compared to existing mainstream object detection al-gorithms,it exhibits significant advantages in both model lightweight and detection accuracy,providing an efficient solution for infrared target detection under scarce data scenarios.关键词
跨模态目标检测/红外图像/数据增强/轻量化/YOLOKey words
cross-modal object detection/infrared images/data augmentation/lightweight/YOLO分类
信息技术与安全科学引用本文复制引用
计锦达,钮赛赛,顾胜宇,汤政,杨志勇,阮洋,陈方言..基于改进YOLO11s的跨模态学习红外目标检测方法[J].飞控与探测,2026,9(1):119-130,12.基金项目
上海航天科技创新基金(SAST2023-054) （SAST2023-054）
