Invariant Representations through Adversarial Forgetting

被引:0
|
作者
Jaiswal, Ayush [1 ]
Moyer, Daniel [1 ]
Ver Steeg, Greg [1 ]
AbdAlmageed, Wael [1 ]
Natarajan, Premkumar [1 ]
机构
[1] Univ Southern Calif, Informat Sci Inst, Los Angeles, CA 90007 USA
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
We propose a novel approach to achieving invariance for deep neural networks in the form of inducing amnesia to unwanted factors of data through a new adversarial forgetting mechanism. We show that the forgetting mechanism serves as an information-bottleneck, which is manipulated by the adversarial training to learn invariance to unwanted factors. Empirical results show that the proposed framework achieves state-of-the-art performance at learning invariance in both nuisance and bias settings on a diverse collection of datasets and tasks.
引用
收藏
页码:4272 / 4279
页数:8
相关论文
共 50 条
  • [1] Invariant Representations without Adversarial Training
    Moyer, Daniel
    Gao, Shuyang
    Brekelmans, Rob
    Steeg, Greg Ver
    Galstyan, Aram
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 31 (NIPS 2018), 2018, 31
  • [2] Learning Signer-Invariant Representations with Adversarial Training
    Ferreira, Pedro M.
    Pernes, Diogo
    Rebelo, Ana
    Cardoso, Jaime S.
    TWELFTH INTERNATIONAL CONFERENCE ON MACHINE VISION (ICMV 2019), 2020, 11433
  • [3] Adversarial Robustness through Disentangled Representations
    Yang, Shuo
    Guo, Tianyu
    Wang, Yunhe
    Xu, Chang
    THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2021, 35 : 3145 - 3153
  • [4] Learning Invariant Representations From EEG via Adversarial Inference
    Ozdenizci, Ozan
    Wang, Ye
    Koike-Akino, Toshiaki
    Erdogmus, Deniz
    IEEE ACCESS, 2020, 8 : 27074 - 27085
  • [5] Improving Word Sense Induction through Adversarial Forgetting of Morphosyntactic Information
    Yavas, Deniz Ekin
    Bernard, Timothée
    Kallmeyer, Laura
    Crabbé, Benoît
    Proceedings of the Annual Meeting of the Association for Computational Linguistics, 2024, : 238 - 251
  • [6] Learning cortical representations through perturbed and adversarial dreaming
    Deperrois, Nicolas
    Petrovici, Mihai A.
    Senn, Walter
    Jordan, Jakob
    Schapiro, Anna C.
    ELIFE, 2022, 11
  • [7] NashAE: Disentangling Representations Through Adversarial Covariance Minimization
    Yeats, Eric
    Liu, Frank
    Womble, David
    Li, Hai
    COMPUTER VISION - ECCV 2022, PT XXVII, 2022, 13687 : 36 - 51
  • [8] Zero-Shot Dense Retrieval with Momentum Adversarial Domain Invariant Representations
    Xin, Ji
    Xiong, Chenyan
    Srinivasan, Ashwin
    Sharma, Ankita
    Jose, Damien
    Bennett, Paul N.
    FINDINGS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2022), 2022, : 4008 - 4020
  • [9] Targeted Forgetting and False Memory Formation in Continual Learners through Adversarial Backdoor Attacks
    Umer, Muhammad
    Dawson, Glenn
    Polikar, Robi
    2020 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2020,
  • [10] Adversarial Invariant Learning
    Ye, Nanyang
    Tang, Jingxuan
    Deng, Huayu
    Zhou, Xiao-Yun
    Li, Qianxiao
    Li, Zhenguo
    Yang, Guang-Zhong
    Zhu, Zhanxing
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 12441 - 12449