郑州大学学报（理学版）2025，Vol.57Issue(3)：42-48,7.DOI:10.13705/j.issn.1671-6841.2023163
基于跨层级注意力学习的RGB-T显著目标检测
RGB-T Salient Object Detection Based on Cross-level Attention Learning
摘要
Abstract
RGB-thermal saliency object detection(RGB-T SOD)aimed to segment common salient re-gions in both visible light images and corresponding thermal infrared images.To address the problem of insufficient utilization of cross-level complementary information among existing methods,a cross-level fea-ture attention learning network(CALNet)was proposed for the RGB-T SOD task.Specifically,the net-work included a cross-level attention learning module(CAL),which used non-local attention to interact cross-level information among multiple modalities and could fully explore global positions and local details across different modalities and levels.In addition,the network also introduced a global information mod-ule(GIM)and a multi-interaction module(MIB),both of which could model and explore multi-type in-formation in a layer-by-layer decoding process for more accurate RGB-T SOD.Extensive experiments on public RGB-T datasets demonstrated that the proposed network achieved excellent performance compared with state-of-the-art algorithms in the field.关键词
多模态/非局部注意力/RGB-T/显著目标检测/特征融合Key words
multimodal/non-local attention/RGB-T/salient object detection/feature fusion分类
计算机与自动化引用本文复制引用
魏明军,魏帅,刘亚志,李辉..基于跨层级注意力学习的RGB-T显著目标检测[J].郑州大学学报（理学版）,2025,57(3):42-48,7.基金项目
河北省高等学校科学技术研究项目(ZD2022102) （ZD2022102）
