Step-Wise Hierarchical Alignment Network for Image-Text Matching

被引:0
|
作者
Ji, Zhong [1 ]
Chen, Kexin [1 ]
Wang, Haoran [1 ]
机构
[1] Tianjin Univ, Sch Elect & Informat Engn, Tianjin, Peoples R China
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Image-text matching plays a central role in bridging the semantic gap between vision and language. The key point to achieve precise visual-semantic alignment lies in capturing the fine-grained cross-modal correspondence between image and text. Most previous methods rely on single-step reasoning to discover the visual-semantic interactions, which lacks the ability of exploiting the multi-level information to locate the hierarchical fine-grained relevance. Different from them, in this work, we propose a step-wise hierarchical alignment network (SHAN) that decomposes image-text matching into multi-step cross-modal reasoning process. Specifically, we first achieve local-to-local alignment at fragment level, following by performing global-to-local and global-to-global alignment at context level sequentially. This progressive alignment strategy supplies our model with more complementary and sufficient semantic clues to understand the hierarchical correlations between image and text. The experimental results on two benchmark datasets demonstrate the superiority of our proposed method.
引用
下载
收藏
页码:765 / 771
页数:7
相关论文
共 50 条
  • [21] Image-text matching algorithm based on multi-level semantic alignment
    Li Y.
    Yao T.
    Zhang L.
    Sun Y.
    Fu H.
    Beijing Hangkong Hangtian Daxue Xuebao/Journal of Beijing University of Aeronautics and Astronautics, 2024, 50 (02): : 551 - 558
  • [22] Prototype local-global alignment network for image-text retrieval
    Meng, Lingtao
    Zhang, Feifei
    Zhang, Xi
    Xu, Changsheng
    INTERNATIONAL JOURNAL OF MULTIMEDIA INFORMATION RETRIEVAL, 2022, 11 (04) : 525 - 538
  • [23] Team HUGE: Image-Text Matching via Hierarchical and Unified Graph Enhancing
    Li, Bo
    Wu, You
    Li, Zhixin
    PROCEEDINGS OF THE 4TH ANNUAL ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA RETRIEVAL, ICMR 2024, 2024, : 704 - 712
  • [24] Unified Adaptive Relevance Distinguishable Attention Network for Image-Text Matching
    Zhang, Kun
    Mao, Zhendong
    Liu, An-An
    Zhang, Yongdong
    IEEE TRANSACTIONS ON MULTIMEDIA, 2023, 25 : 1320 - 1332
  • [25] Alignment of Image-Text and Video-Text Datasets
    Ozkose, Yunus Emre
    Gokce, Zeynep
    Duygulu, Pinar
    2023 31ST SIGNAL PROCESSING AND COMMUNICATIONS APPLICATIONS CONFERENCE, SIU, 2023,
  • [26] Multi-scale motivated neural network for image-text matching
    Qin, Xueyang
    Li, Lishuang
    Pang, Guangyao
    MULTIMEDIA TOOLS AND APPLICATIONS, 2023, 83 (2) : 4383 - 4407
  • [27] Dual-View Semantic Inference Network for image-text matching
    Wu, Chunlei
    Wu, Jie
    Cao, Haiwen
    Wei, Yiwei
    Wang, Leiquan
    NEUROCOMPUTING, 2021, 426 : 47 - 57
  • [28] Cross-modal Semantically Augmented Network for Image-text Matching
    Yao, Tao
    Li, Yiru
    Li, Ying
    Zhu, Yingying
    Wang, Gang
    Yue, Jun
    ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS, 2024, 20 (04)
  • [29] Global-Guided Asymmetric Attention Network for Image-Text Matching
    Wu, Dongqing
    Li, Huihui
    Tang, Yinge
    Guo, Lei
    Liu, Hang
    NEUROCOMPUTING, 2022, 481 : 77 - 90
  • [30] Multi-scale motivated neural network for image-text matching
    Xueyang Qin
    Lishuang Li
    Guangyao Pang
    Multimedia Tools and Applications, 2024, 83 : 4383 - 4407