Exploring Explicitly Disentangled Features for Domain Generalization
文献类型:期刊论文
作者 | Li, Jingwei1,2; Li, Yuan1,2; Wang, Huanjie1,2; Liu, Chengbao2; Tan, Jie1,2 |
刊名 | IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY |
出版日期 | 2023-11-01 |
卷号 | 33期号:11页码:6360-6373 |
ISSN号 | 1051-8215 |
关键词 | Domain generalization feature disentanglement Fourier transform data augmentation |
DOI | 10.1109/TCSVT.2023.3269534 |
通讯作者 | Tan, Jie(tan.jie@tom.com) |
英文摘要 | Domain generalization (DG) is a challenging task that aims to train a robust model with only labeled source data and can generalize well on unseen target data. The domain gap between the source and target data may degrade the performance. A plethora of methods resort to obtaining domain-invariant features to overcome the difficulties. However, these methods require sophisticated network designs or training strategies, causing inefficiency and complexity. In this paper, we first analyze and reclassify the features into two categories, i.e., implicitly disentangled ones and explicitly disentangled counterparts. Since we aim to design a generic algorithm for DG to alleviate the problems mentioned above, we focus more on the explicitly disentangled features due to their simplicity and interpretability. We find out that the shape features of images are simple and elegant choices based on our analysis. We extract the shape features from two aspects. In the aspect of networks, we propose Multi-Scale Amplitude Mixing (MSAM) to strengthen shape features at different layers of the network by Fourier transform. In the aspect of inputs, we propose a new data augmentation method called Random Shape Warping (RSW) to facilitate the model to concentrate more on the global structures of the objects. RSW randomly distorts the local parts of the images and keeps the global structures unchanged, which can further improve the robustness of the model. Our methods are simple yet efficient and can be conveniently used as plug-and-play modules. They can outperform state-of-the-art (SOTA) methods without bells and whistles. |
资助项目 | National Key Research and Development Program of China[2022YFB3304602] ; National Natural Science Foundation of China[62003344] |
WOS研究方向 | Engineering |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:001093434100012 |
资助机构 | National Key Research and Development Program of China ; National Natural Science Foundation of China |
源URL | [http://ir.ia.ac.cn/handle/173211/54433] |
专题 | 中科院工业视觉智能装备工程实验室 |
通讯作者 | Tan, Jie |
作者单位 | 1.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 101408, Peoples R China 2.Chinese Acad Sci, Inst Automat, Beijing 100080, Peoples R China |
推荐引用方式 GB/T 7714 | Li, Jingwei,Li, Yuan,Wang, Huanjie,et al. Exploring Explicitly Disentangled Features for Domain Generalization[J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,2023,33(11):6360-6373. |
APA | Li, Jingwei,Li, Yuan,Wang, Huanjie,Liu, Chengbao,&Tan, Jie.(2023).Exploring Explicitly Disentangled Features for Domain Generalization.IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,33(11),6360-6373. |
MLA | Li, Jingwei,et al."Exploring Explicitly Disentangled Features for Domain Generalization".IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 33.11(2023):6360-6373. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。