Faster Meta Update Strategy for Noise-Robust Deep Learning

被引:37
|
作者
Xu, Youjiang [1 ]
Zhu, Linchao [2 ]
Jiang, Lu [3 ]
Yang, Yi [2 ]
机构
[1] Baidu Res, Beijing, Peoples R China
[2] Univ Technol Sydney, ReLER, Sydney, NSW, Australia
[3] Google Res, Mountain View, CA USA
关键词
D O I
10.1109/CVPR46437.2021.00021
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
It has been shown that deep neural networks are prone to overfitting on biased training data. Towards addressing this issue, meta-learning employs a meta model for correcting the training bias. Despite the promising performances, super slow training is currently the bottleneck in the meta learning approaches. In this paper, we introduce a novel Faster Meta Update Strategy (FaMUS) to replace the most expensive step in the meta gradient computation with a faster layer-wise approximation. We empirically find that FaMUS yields not only a reasonably accurate but also a low-variance approximation of the meta gradient. We conduct extensive experiments to verify the proposed method on two tasks. We show our method is able to save two-thirds of the training time while still maintaining the comparable or achieving even better generalization performance. In particular, our method achieves the state-of-the-art performance on both synthetic and realistic noisy labels, and obtains promising performance on long-tailed recognition on standard benchmarks.
引用
收藏
页码:144 / 153
页数:10
相关论文
共 50 条
  • [1] Training Noise-Robust Deep Neural Networks via Meta-Learning
    Wang, Zhen
    Hu, Guosheng
    Hu, Qinghua
    2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2020, : 4523 - 4532
  • [2] Sparse Representations in Deep Learning for Noise-Robust Digit Classification
    Ghifary, Muhammad
    Kleijn, W. Bastiaan
    Zhang, Mengjie
    PROCEEDINGS OF 2013 28TH INTERNATIONAL CONFERENCE ON IMAGE AND VISION COMPUTING NEW ZEALAND (IVCNZ 2013), 2013, : 340 - 345
  • [3] Optimizing Deep Learning for Efficient and Noise-Robust License Plate Detection and Recognition
    Shim, Seong-O
    Imtiaz, Romil
    Habibullah, Safa
    Alshdadi, Abdulrahman A.
    INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS, 2024, 15 (04) : 595 - 607
  • [4] Enhancing traffic monitoring with noise-robust distributed acoustic sensing and deep learning
    Wang, Zheng
    Zhang, Taiyin
    Chen, Huiliang
    Zhang, Cheng-Cheng
    Shi, Bin
    JOURNAL OF APPLIED GEOPHYSICS, 2025, 233
  • [5] Noise-Robust Deep Learning Model for Emotion Classification Using Facial Expressions
    Oh, Seungjun
    Kim, Dong-Keun
    IEEE ACCESS, 2024, 12 : 143074 - 143089
  • [6] Noise-Robust Sampling for Collaborative Metric Learning
    Matsui, Ryo
    Yaginuma, Suguru
    Naito, Taketo
    Nakata, Kazuhide
    REVIEW OF SOCIONETWORK STRATEGIES, 2022, 16 (02): : 307 - 332
  • [7] Software Entity Recognition with Noise-Robust Learning
    Tai Nguyen
    Di, Yifeng
    Lee, Joohan
    Chen, Muhao
    Zhang, Tianyi
    2023 38TH IEEE/ACM INTERNATIONAL CONFERENCE ON AUTOMATED SOFTWARE ENGINEERING, ASE, 2023, : 484 - 496
  • [8] A CASCADED NOISE-ROBUST DEEP CNN FOR FACE RECOGNITION
    Meng, Xiangbang
    Yan, Yan
    Chen, Si
    Wang, Hanzi
    2019 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP), 2019, : 3487 - 3491
  • [9] Noise-robust Deep Cross-Modal Hashing
    Wang, Runmin
    Yu, Guoxian
    Zhang, Hong
    Guo, Maozu
    Cui, Lizhen
    Zhang, Xiangliang
    INFORMATION SCIENCES, 2021, 581 : 136 - 154
  • [10] Noise-Robust Federated Learning With Model Heterogeneous Clients
    Fang, Xiuwen
    Ye, Mang
    IEEE TRANSACTIONS ON MOBILE COMPUTING, 2025, 24 (05) : 4053 - 4071