Adaptive Graph Learning With Semantic Promotability for Domain Adaptation

被引:0
作者
Zheng, Zefeng [1 ]
Teng, Shaohua [1 ,2 ]
Teng, Luyao [3 ]
Zhang, Wei [1 ]
Wu, Naiqi [4 ,5 ]
机构
[1] Guangdong Univ Technol, Sch Comp Sci & Technol, Guangzhou 510006, Peoples R China
[2] Guangdong Univ ofTechnol, Sch Adv Mfg, Guangzhou 510006, Peoples R China
[3] Sch Informat Engn, Guangzhou Panyu Polytech, Guangzhou 511483, Peoples R China
[4] Macau Univ Sci & Technol, Inst Syst Engn, Macau 999078, Peoples R China
[5] Macau Univ Sci & Technol, Collaborat Lab Intelligent Sci & Syst, Macau 999078, Peoples R China
基金
中国国家自然科学基金;
关键词
Adaptive graph learning; domain adaptation; implicit semantics preservation; semantically promotable sample enhancement; variant semantic and geometrical component learning;
D O I
10.1109/TPAMI.2024.3507534
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Domain Adaptation (DA) is used to reduce cross-domain differences between the labeled source and unlabeled target domains. As the existing semantic-based DA approaches mainly focus on extracting consistent knowledge under semantic guidance, they may fail in acquiring: (a) personalized knowledge between intra-class samples and (b) local knowledge of neighbor samples from different categories. Hence, a multi-semantic-granularity and target-sample oriented approach, called Adaptive Graph Learning with Semantic Promotability (AGLSP), is proposed, which consists of three parts: (a) Adaptive Graph Embedding with Semantic Guidance (AGE-SG) that adaptively estimates the promotability of target samples and learns variant semantic and geometrical components from the source and those semantically promotable target samples; (b) Semantically Promotable Sample Enhancement (SPSE) that further increases the discriminability and adaptability of tag granularity by mining the features of intra-class source and semantically promotable target samples with multi-granularities; and (c) Adaptive Graph Learning with Implicit Semantic Preservation (AGL-ISP) that forms the tag granularity by extracting commonalities between the source and those semantically non-promotable target samples. As AGLSP learns more semantics from the two domains, more cross-domain knowledge is transferred. Mathematical proofs and extensive experiments on seven datasets demonstrate the performance of AGLSP.
引用
收藏
页码:1747 / 1763
页数:17
相关论文
共 50 条
[41]   Correntropy-Induced Wasserstein GCN: Learning Graph Embedding via Domain Adaptation [J].
Wang, Wei ;
Zhang, Gaowei ;
Han, Hongyong ;
Zhang, Chi .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2023, 32 :3980-3993
[42]   Source-Free Progressive Graph Learning for Open-Set Domain Adaptation [J].
Luo, Yadan ;
Wang, Zijian ;
Chen, Zhuoxiao ;
Huang, Zi ;
Baktashmotlagh, Mahsa .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2023, 45 (09) :11240-11255
[43]   Rethinking unsupervised domain adaptation for semantic segmentation [J].
Wang, Zhijie ;
Suganuma, Masanori ;
Okatani, Takayuki .
PATTERN RECOGNITION LETTERS, 2024, 186 :119-125
[44]   Unsupervised Domain Adaptation for Referring Semantic Segmentation [J].
Shi, Haonan ;
Pan, Wenwen ;
Zhao, Zhou ;
Zhang, Mingmin ;
Wu, Fei .
PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2023, 2023, :5807-5818
[45]   Informative pairs mining based adaptive metric learning for adversarial domain adaptation [J].
Wang, Mengzhu ;
Li, Paul ;
Shen, Li ;
Wang, Ye ;
Wang, Shanshan ;
Wang, Wei ;
Zhang, Xiang ;
Chen, Junyang ;
Luo, Zhigang .
NEURAL NETWORKS, 2022, 151 :238-249
[46]   Cross-Domain Transformer with Adaptive Thresholding for Domain Adaptive Semantic Segmentation [J].
Liu, Quansheng ;
Wang, Lei ;
Jun, Yu ;
Gao, Fang .
ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING, ICANN 2023, PT VIII, 2023, 14261 :147-159
[47]   STRUCTURAL DOMAIN ADAPTATION WITH LATENT GRAPH ALIGNMENT [J].
Zhang, Yue ;
Miao, Shun ;
Liao, Rui .
2018 25TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP), 2018, :3753-3757
[48]   Adaptive Component Embedding for Domain Adaptation [J].
Jing, Mengmeng ;
Zhao, Jidong ;
Li, Jingjing ;
Zhu, Lei ;
Yang, Yang ;
Shen, Heng Tao .
IEEE TRANSACTIONS ON CYBERNETICS, 2021, 51 (07) :3390-3403
[49]   Semantic Aware Answer Sentence Selection using Self-Learning based Domain Adaptation [J].
Sarkar, Rajdeep ;
Dutta, Sourav ;
Assem, Haytham ;
Arcan, Mihael ;
McCrae, John .
PROCEEDINGS OF THE 28TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, KDD 2022, 2022, :3849-3857
[50]   Domain Adaptation Semantic Segmentation for Urban Scene Combining Self-ensembling and Adversarial Learning [J].
Zhang G. ;
Lu F. ;
Long B. ;
Miao J. .
Moshi Shibie yu Rengong Zhineng/Pattern Recognition and Artificial Intelligence, 2021, 34 (01) :58-67