Bi-Modality Individual-Aware Prompt Tuning for Visual-Language Model

doi:10.1109/TPAMI.2025.3557780

	Bi-Modality Individual-Aware Prompt Tuning for Visual-Language Model
	Yao, Hantao 1; Zhang, Rui 2; Lyu, Huaihai 3,4; Zhang, Yongdong 1; Xu, Changsheng 3,4
	2025-08-01
发表期刊	IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
ISSN	0162-8828
卷号	47 期号:8 页码:6352-6368
摘要	Prompt tuning is a valuable technique for adapting visual language models (VLMs) to different downstream tasks, such as domain generalization and learning from a few examples. Previous methods have utilized Context Optimization approaches to deduce domain-shared or cross-modality prompt tokens, which enhance generalization and discriminative ability in textual or visual contexts. However, these prompt tokens, inferred from training data, cannot adapt perfectly to the distribution of the test dataset. This work introduces a novel approach called Bi-modality Individual-aware Prompt Tuning (BIP) by explicitly incorporating the individual's essential prior knowledge into the learnable prompt to enhance their discriminability and generalization. The critical insight of BIP involves applying the Textual Knowledge Embedding (TKE) and Visual Knowledge Embedding (VKE) models to project the class-aware textual essential knowledge and the instance-aware essential knowledge into the class-aware prompt and instance-aware prompt, referred to as Textual-level Class-aware Prompt tuning (TCP) and Visual-level Instance-aware Prompt tuning (VIP). On the one hand, TCP integrates the generated class-aware prompts into the Text Encoder to produce a dynamic class-aware classifier to improve generalization on unseen domains. On the other hand, VIP uses the instance-aware prompt to generate the dynamic visual embedding of each instance, thereby enhancing the discriminative capability of visual embedding. Comprehensive evaluations demonstrate that BIP can be used as a plug-and-play module easily integrated with existing methods and achieves superior performance on 15 benchmarks across four tasks.
关键词	Tuning Visualization Training Adaptation models Hands Feature extraction Data models Artificial intelligence Transformers Data mining Prompt tuning individual-aware prompt tuning bi-modality prompt tuning visual-language model
DOI	10.1109/TPAMI.2025.3557780
收录类别	SCI
语种	英语
资助项目	National Science and Technology[2021ZD0112202] ; National Natural Science Foundation of China[62376268] ; National Natural Science Foundation of China[U23A20387] ; National Natural Science Foundation of China[U21B2044] ; National Natural Science Foundation of China[62121002]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:001522958700031
出版者	IEEE COMPUTER SOC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/41767
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Yao, Hantao
作者单位	1.Univ Sci & Technol China, Sch Informat Sci & Technol, Hefei 230026, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, State Key Lab Proc, Beijing 100190, Peoples R China 3.Chinese Acad Sci, Inst Automat, State Key Lab Multimodal Artificial Intelligence S, Beijing 100190, Peoples R China 4.Univ Chinese Acad Sci, Beijing 100049, Peoples R China
推荐引用方式 GB/T 7714	Yao, Hantao,Zhang, Rui,Lyu, Huaihai,et al. Bi-Modality Individual-Aware Prompt Tuning for Visual-Language Model[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,2025,47(8):6352-6368.
APA	Yao, Hantao,Zhang, Rui,Lyu, Huaihai,Zhang, Yongdong,&Xu, Changsheng.(2025).Bi-Modality Individual-Aware Prompt Tuning for Visual-Language Model.IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,47(8),6352-6368.
MLA	Yao, Hantao,et al."Bi-Modality Individual-Aware Prompt Tuning for Visual-Language Model".IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 47.8(2025):6352-6368.