重庆邮电大学学报（自然科学版）2025，Vol.37Issue(1)：67-75,9.DOI:10.3979/j.issn.1673-825X.202312210427
基于阅读策略和语义对齐的图文匹配方法
Image-text matching based on reading strategy and semantic alignment
摘要
Abstract
To address the image-text matching task in the cross-media computing domain,this paper proposes a reading strategy and semantic alignment network(RSAN).A region feature enhancement module based on transformer and bidirec-tional gated recurrent units(Bi-GRU)is designed to generate image region features with semantic relationships,improving the accuracy of semantic alignment.A reading module containing an overview branch and a close-reading branch is designed to aggregate global and local alignments for learning more accurate matching scores.Comprehensive experiments conducted on the Flickr30K and MS-COCO datasets show that the RSAN model outperforms existing baseline models in both accuracy and efficiency.关键词
图文匹配/特征增强/语义对齐/相似度计算Key words
image-text matching/feature enhancement/semantic alignment/similarity computation分类
信息技术与安全科学引用本文复制引用
甘凤梅,夏英..基于阅读策略和语义对齐的图文匹配方法[J].重庆邮电大学学报（自然科学版）,2025,37(1):67-75,9.基金项目
国家自然科学基金项目(41971365) （41971365）
重庆市教委重点合作项目(HZ2021008) （HZ2021008）
文化和旅游部重点实验室资助项目(E020H2023005)National Natural Science Foundation of China(41971365) （E020H2023005）
Chongqing Municipal Education Commission Key Co-operation Projects(HZ2021008) （HZ2021008）
Funded Project of Key Laboratory under Ministry of Culture and Tourism(E020H2023005) （E020H2023005）
