Equal Loss: A Simple Loss Function for Noise Robust Learning

Lei Cui; Huan Peng; Yangguang Li; Chuming Li; Xingrun Xing

ICASSP 2022accepted0 citations

Equal Loss: A Simple Loss Function for Noise Robust Learning

Lei Cui, Huan Peng, Yangguang Li, Chuming Li, Xingrun Xing

Abstract

Training accurate deep neural networks in the presence of noisy labels is an important task. Though a number of approaches have been proposed for learning with noisy labels, many open issues remain. In this paper, we show that DNN learning with Cross Entropy is not robust to label noise and exhibits imbalance between the gradient of clean and noisy samples. We propose a new loss function, Equal Loss (EL), boosting DNN with a relaxed target probability and balanced gradient density. Both theoretical analysis and experiments on a range of benchmarks and real-world datasets show that EL outperforms state-of-the-art methods.

BibTeX

@inproceedings{icassp2022_equallossasimple,
  title = {Equal Loss: A Simple Loss Function for Noise Robust Learning},
  author = {Lei Cui and Huan Peng and Yangguang Li and Chuming Li and Xingrun Xing},
  booktitle = {ICASSP 2022},
  year = {2022}
}