Dual-Branch Task Residual Enhancement with Parameter-Free Attention for Zero-Shot Multi-label Image Recognition

Shizhou Zhang, Kairui Dang, De Cheng, Yinghui Xing, Qirui Wu, Dexuan Kong, Yanning Zhang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Zero-shot multi-label image recognition involves the task of recognizing multi-label images while “zero” visual information has been input into the model during training. Recently, with the emergence of large pre-trained vision-language model, the visual and semantic features can be well aligned after being trained with billions of image-text pairs collected from the internet. In this paper, by utilizing the pre-trained CLIP model, we propose a dual-branch task residual enhancement with parameter-free attention module that enhances interaction of inter-modal information to tackle the problem of multi-label image recognition. The method employs a dual-branch structure, including global and local branches. The local branch mitigates global feature dominance, improving image content understanding ability of local regions. Our method shows superiority in zero-shot multi-label learning on VOC2007, MS-COCO, and NUS-WIDE datasets, surpassing the state-of-the-art methods. Additionally, it also has excellent performance in partial label settings. Code is available in the supplementary materials.

Original languageEnglish
Title of host publicationPattern Recognition - 27th International Conference, ICPR 2024, Proceedings
EditorsApostolos Antonacopoulos, Subhasis Chaudhuri, Rama Chellappa, Cheng-Lin Liu, Saumik Bhattacharya, Umapada Pal
PublisherSpringer Science and Business Media Deutschland GmbH
Pages160-171
Number of pages12
ISBN (Print)9783031783111
DOIs
StatePublished - 2025
Event27th International Conference on Pattern Recognition, ICPR 2024 - Kolkata, India
Duration: 1 Dec 20245 Dec 2024

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume15322 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference27th International Conference on Pattern Recognition, ICPR 2024
Country/TerritoryIndia
CityKolkata
Period1/12/245/12/24

Keywords

  • Dual-branch
  • Multi-label
  • Task Residual
  • Zero-shot

Fingerprint

Dive into the research topics of 'Dual-Branch Task Residual Enhancement with Parameter-Free Attention for Zero-Shot Multi-label Image Recognition'. Together they form a unique fingerprint.

Cite this