Improving Inconspicuous Attributes Modeling for Person Search by Language

doi:10.1109/TIP.2023.3285426

CORC > 自动化研究所 > 中国科学院自动化研究所 > 多模态人工智能系统全国重点实验室

	Improving Inconspicuous Attributes Modeling for Person Search by Language
	Niu, Kai 2,3; Huang, Tao 2,3; Huang, Linjiang 4; Wang, Liang 1; Zhang, Yanning 3
刊名	IEEE TRANSACTIONS ON IMAGE PROCESSING
	2023
卷号	32 页码:3429-3441
关键词	Person search by language cross-modal retrieval smart video surveillance
ISSN号	1057-7149
DOI	10.1109/TIP.2023.3285426
通讯作者	Niu, Kai(kai.niu@nwpu.edu.cn) ; Huang, Linjiang(ljhuang524@gmail.com)
英文摘要	Person search by language aims to retrieve the interested pedestrian images based on natural language sentences. Although great efforts have been made to address the cross-modal heterogeneity, most of the current solutions suffer from only capturing salient attributes while ignoring inconspicuous ones, being weak in distinguishing very similar pedestrians. In this work, we propose the Adaptive Salient Attribute Mask Network (ASAMN) to adaptively mask the salient attributes for cross-modal alignments, and therefore induce the model to simultaneously focus on inconspicuous attributes. Specifically, we consider the uni-modal and cross-modal relations for masking salient attributes in the Uni-modal Salient Attribute Mask (USAM) and Cross-modal Salient Attribute Mask (CSAM) modules, respectively. Then the Attribute Modeling Balance (AMB) module is presented to randomly select a proportion of masked features for cross-modal alignments, ensuring the balance of modeling capacity of both salient attributes and inconspicuous ones. Extensive experiments and analyses have been carried out to validate the effectiveness and generalization capacity of our proposed ASAMN method, and we have obtained the state-of-the-art retrieval performance on the widely-used CUHK-PEDES and ICFG-PEDES benchmarks.
资助项目	National Natural Science Foundation of China[62101451] ; National Natural Science Foundation of China[U19B2037] ; Guangdong Basic and Applied Basic Research Foundation[2023A1515011427] ; Fundamental Research Funds for the Central Universities[D5000210733]
WOS关键词	VIDEO ; IMAGE
WOS研究方向	Computer Science ; Engineering
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:001017283200006
资助机构	National Natural Science Foundation of China ; Guangdong Basic and Applied Basic Research Foundation ; Fundamental Research Funds for the Central Universities
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/53678]
专题	多模态人工智能系统全国重点实验室
通讯作者	Niu, Kai; Huang, Linjiang
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.Northwestern Polytech Univ Shenzhen, Res & Dev Inst, Shenzhen 518063, Peoples R China 3.Northwestern Polytech Univ, Sch Comp Sci, Natl Engn Lab Integrated Aerosp Ground Ocean Big D, Xian 710129, Peoples R China 4.Chinese Univ Hong Kong, Multimedia Lab, Hong Kong, Peoples R China
推荐引用方式 GB/T 7714	Niu, Kai,Huang, Tao,Huang, Linjiang,et al. Improving Inconspicuous Attributes Modeling for Person Search by Language[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2023,32:3429-3441.
APA	Niu, Kai,Huang, Tao,Huang, Linjiang,Wang, Liang,&Zhang, Yanning.(2023).Improving Inconspicuous Attributes Modeling for Person Search by Language.IEEE TRANSACTIONS ON IMAGE PROCESSING,32,3429-3441.
MLA	Niu, Kai,et al."Improving Inconspicuous Attributes Modeling for Person Search by Language".IEEE TRANSACTIONS ON IMAGE PROCESSING 32(2023):3429-3441.