Class-Wise Denoising for Robust Learning under Label Noise

Chen Gong*, Yongliang Ding, Bo Han, Gang Niu, Jian Yang*, Jane J. You, Dacheng Tao, Masashi Sugiyama

*Corresponding author for this work

Research output: Contribution to journalJournal articlepeer-review

8 Citations (Scopus)


Label noise is ubiquitous in many real-world scenarios which often misleads training algorithm and brings about the degraded classification performance. Therefore, many approaches have been proposed to correct the loss function given corrupted labels to combat such label noise. Among them, a trend of works achieve this goal by unbiasedly estimating the data centroid, which plays an important role in constructing an unbiased risk estimator for minimization. However, they usually handle the noisy labels in different classes all at once, so the local information inherited by each class is ignored which often leads to unsatisfactory performance. To address this defect, this paper presents a novel robust learning algorithm dubbed 'Class-Wise Denoising' (CWD), which tackles the noisy labels in a class-wise way to ease the entire noise correction task. Specifically, two virtual auxiliary sets are respectively constructed by presuming that the positive and negative labels in the training set are clean, so the original false-negative labels and false-positive ones are tackled separately. As a result, an improved centroid estimator can be designed which helps to yield more accurate risk estimator. Theoretically, we prove that: 1) the variance in centroid estimation can often be reduced by our CWD when compared with existing methods with unbiased centroid estimator; and 2) the performance of CWD trained on the noisy set will converge to that of the optimal classifier trained on the clean set with a convergence rate O(1n) where n is the number of the training examples. These sound theoretical properties critically enable our CWD to produce the improved classification performance under label noise, which is also demonstrated by the comparisons with ten representative state-of-the-art methods on a variety of benchmark datasets.

Original languageEnglish
Pages (from-to)2835-2848
Number of pages14
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Issue number3
Early online date30 May 2022
Publication statusPublished - 1 Mar 2023

Scopus Subject Areas

  • Software
  • Artificial Intelligence
  • Applied Mathematics
  • Computer Vision and Pattern Recognition
  • Computational Theory and Mathematics

User-Defined Keywords

  • Centroid estimation
  • Label noise
  • Unbiasedness
  • Variance reduction


Dive into the research topics of 'Class-Wise Denoising for Robust Learning under Label Noise'. Together they form a unique fingerprint.

Cite this