EUVIMEDEUVIMED entdecken
Crossref

A Time-Consistency Curriculum for Learning From Instance-Dependent Noisy Labels

Songhua Wu, Tianyi Zhou, Yuxuan Du, Jun Yu, Bo Han, Tongliang Liu

IEEE Transactions on Pattern Analysis and Machine Intelligence · 2024-07 · 46 (7) · 4830-4842

DOI: 10.1109/tpami.2024.3360623 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024-07. Verfasst von Songhua Wu und 5 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 64 Literaturangaben. Gefördert von: Australian Research Council Project, RIKEN Collaborative Research Fund, Australian Research Council.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Many machine learning algorithms are known to be fragile on simple instance-independent noisy labels. However, noisy labels in real-world data are more devastating since they are produced by more complicated mechanisms in an instance-dependent manner. In this paper, we target this practical challenge of Instance-Dependent Noisy Labels by jointly training (1) a model reversely engineering the noise generating mechanism, which produces an instance-dependent mapping between the clean label posterior and the observed noisy label and (2) a robust classifier that produces clean label posteriors. Compared to previous methods, the former model is novel and enables end-to-end learning of the latter directly from noisy labels. An extensive empirical study indicates that the time-consistency of data is critical to the success of training both models and motivates us to develop a curriculum selecting training data based on their dynamics on the two models' outputs over the course of training. We show that the curriculum-selected data provide both clean labels and high-quality input-output pairs for training the two models. Therefore, it leads to promising and robust classification performance even in notably challenging settings of instance-dependent noisy labels where many SoTA methods could easily fail. Extensive experimental comparisons and ablation studies further demonstrate the advantages and significance of the time-consistency curriculum in learning from instance-dependent noisy labels on multiple benchmark datasets.

Förderung

  • Australian Research Council Project
  • RIKEN Collaborative Research Fund
  • Australian Research Council
  • National Natural Science Foundation of China
  • National Aviation Science Foundation
  • CAAI-Huawei MindSpore Open Fund
  • Anhui Province Key Research and Development Program
  • Dreams Foundation of Jianghuai Advance Technology Center

Zitieren

APA

Wu, S., Zhou, T., Du, Y., Yu, J., Han, B., & Liu, T. (2024). A Time-Consistency Curriculum for Learning From Instance-Dependent Noisy Labels. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(7), 4830-4842. https://doi.org/10.1109/tpami.2024.3360623

Vancouver

Wu S, Zhou T, Du Y, Yu J, Han B, Liu T. A Time-Consistency Curriculum for Learning From Instance-Dependent Noisy Labels. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2024;46(7):4830-4842. doi:10.1109/tpami.2024.3360623

BibTeX
@article{Wu2024,
  author = {Wu, Songhua and Zhou, Tianyi and Du, Yuxuan and Yu, Jun and Han, Bo and Liu, Tongliang},
  title = {A Time-Consistency Curriculum for Learning From Instance-Dependent Noisy Labels},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2024},
  volume = {46},
  number = {7},
  pages = {4830--4842},
  doi = {10.1109/tpami.2024.3360623}
}
Kostenlos starten →