Domain Adaptation with Adversarial Training on Penultimate Activations

被引:0
|
作者
Sun, Tao [1 ]
Lu, Cheng [2 ]
Ling, Haibin [1 ]
机构
[1] SUNY Stony Brook, Stony Brook, NY 11794 USA
[2] XPeng Motors, Guangzhou, Peoples R China
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Enhancing model prediction confidence on target data is an important objective in Unsupervised Domain Adaptation (UDA). In this paper, we explore adversarial training on penultimate activations, i.e., input features of the final linear classification layer. We show that this strategy is more efficient and better correlated with the objective of boosting prediction confidence than adversarial training on input images or intermediate features, as used in previous works. Furthermore, with activation normalization commonly used in domain adaptation to reduce domain gap, we derive two variants and systematically analyze the effects of normalization on our adversarial training. This is illustrated both in theory and through empirical analysis on real adaptation tasks. Extensive experiments are conducted on popular UDA benchmarks under both standard setting and source-data free setting. The results validate that our method achieves the best scores against previous arts. Code is available at https://github.com/tsun/APA.
引用
收藏
页码:9935 / 9943
页数:9
相关论文
共 50 条
  • [31] ADVERSARIAL DOMAIN SEPARATION AND ADAPTATION
    Tsai, Jen-Chieh
    Chien, Jen-Tzung
    2017 IEEE 27TH INTERNATIONAL WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING, 2017,
  • [32] Joint Adversarial Domain Adaptation
    Li, Shuang
    Liu, Chi Harold
    Xie, Binhui
    Su, Limin
    Ding, Zhengming
    Huang, Gao
    PROCEEDINGS OF THE 27TH ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA (MM'19), 2019, : 729 - 737
  • [33] Consensus Adversarial Domain Adaptation
    Zou, Han
    Zhou, Yuxun
    Yang, Jianfei
    Liu, Huihan
    Das, Hari Prasanna
    Spanos, Costas J.
    THIRTY-THIRD AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FIRST INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / NINTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2019, : 5997 - 6004
  • [34] ADAPT: Adversarial Domain Adaptation with Purifier Training for Cross-Domain Credit Risk Forecasting
    Zeng, Guanxiong
    Chi, Jianfeng
    Ma, Rui
    Feng, Jinghua
    Ao, Xiang
    Yang, Hao
    DATABASE SYSTEMS FOR ADVANCED APPLICATIONS, DASFAA 2022, PT I, 2022, : 353 - 369
  • [35] Toward Enhanced Adversarial Robustness Generalization in Object Detection: Feature Disentangled Domain Adaptation for Adversarial Training
    Jung, Yoojin
    Song, Byung Cheol
    IEEE ACCESS, 2024, 12 : 179065 - 179076
  • [36] When Adversarial Training Meets Prompt Tuning: Adversarial Dual Prompt Tuning for Unsupervised Domain Adaptation
    Cui, Chaoran
    Liu, Ziyi
    Gong, Shuai
    Zhu, Lei
    Zhang, Chunyun
    Liu, Hui
    IEEE TRANSACTIONS ON IMAGE PROCESSING, 2025, 34 : 1427 - 1440
  • [37] An unsupervised domain adaptation model based on dual-module adversarial training
    Yang, Yiju
    Zhang, Tianxiao
    Li, Guanyu
    Kim, Taejoon
    Wang, Guanghui
    NEUROCOMPUTING, 2022, 475 : 102 - 111
  • [38] Unsupervised Reverse Domain Adaptation for Synthetic Medical Images via Adversarial Training
    Mahmood, Faisal
    Chen, Richard
    Durr, Nicholas J.
    IEEE TRANSACTIONS ON MEDICAL IMAGING, 2018, 37 (12) : 2572 - 2581
  • [39] ATPL: Mutually enhanced adversarial training and pseudo labeling for unsupervised domain adaptation
    Yi, Chang'an
    Chen, Haotian
    Xu, Yonghui
    Liu, Yong
    Jiang, Lei
    Tan, Haishu
    KNOWLEDGE-BASED SYSTEMS, 2022, 250
  • [40] Adversarially Robust Source-free Domain Adaptation with Relaxed Adversarial Training
    Xiao, Yao
    Wei, Pengxu
    Liu, Cong
    Lin, Liang
    2023 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, ICME, 2023, : 2681 - 2686