Gradient-aware domain-invariant learning for domain generalization

doi:10.1007/s00530-024-01613-4

	Gradient-aware domain-invariant learning for domain generalization
	Hou, Feng 1,2; Zhang, Yao 3; Liu, Yang 1,2; Yuan, Jin 4; Zhong, Cheng 3; Zhang, Yang 3; Shi, Zhongchao 3; Fan, Jianping 3; He, Zhiqiang 1,2,5
	2025-02-01
发表期刊	MULTIMEDIA SYSTEMS
ISSN	0942-4962
卷号	31 期号:1 页码:15
摘要	In realistic scenarios, the effectiveness of Deep Neural Networks is hindered by domain shift, where discrepancies between training (source) and testing (target) domains lead to poor generalization on previously unseen data. The Domain Generalization (DG) paradigm addresses this challenge by developing a general model that relies solely on source domains, aiming for robust performance in unknown domains. Despite the progress of prior augmentation-based methods by introducing more diversity based on the known distribution, DG still suffers from overfitting due to limited domain-specific information. Therefore, unlike prior DG methods that treat all parameters equally, we propose a Gradient-Aware Domain-Invariant Learning mechanism that adaptively recognizes and emphasizes domain-invariant parameters. Specifically, two novel models named Domain Decoupling and Combination and Domain-Invariance-Guided Backpropagation (DIGB) are introduced to first generate contrastive samples with the same domain-invariant features and then selectively prioritize parameters with unified optimization directions across contrastive sample pairs to enhance domain robustness. Additionally, a sparse version of DIGB achieves a trade-off between performance and efficiency. Our extensive experiments on various domain generalization benchmarks demonstrate that our proposed method achieves state-of-the-art performance with strong generalization capabilities.
关键词	Domain shift Domain generalization Domain-invariant parameters Sparse
DOI	10.1007/s00530-024-01613-4
收录类别	SCI
语种	英语
WOS研究方向	Computer Science
WOS类目	Computer Science, Information Systems ; Computer Science, Theory & Methods
WOS记录号	WOS:001389241300003
出版者	SPRINGER
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40773
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	He, Zhiqiang
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China 2.Univ Chinese Acad Sci, Beijing, Peoples R China 3.Lenovo Res, AI Lab, Beijing, Peoples R China 4.Southeast Univ, Nanjing, Peoples R China 5.Lenovo Ltd, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Hou, Feng,Zhang, Yao,Liu, Yang,et al. Gradient-aware domain-invariant learning for domain generalization[J]. MULTIMEDIA SYSTEMS,2025,31(1):15.
APA	Hou, Feng.,Zhang, Yao.,Liu, Yang.,Yuan, Jin.,Zhong, Cheng.,...&He, Zhiqiang.(2025).Gradient-aware domain-invariant learning for domain generalization.MULTIMEDIA SYSTEMS,31(1),15.
MLA	Hou, Feng,et al."Gradient-aware domain-invariant learning for domain generalization".MULTIMEDIA SYSTEMS 31.1(2025):15.