Conference Proceedings

Symmetric cross entropy for robust learning with noisy labels

Y Wang, X Ma, Z Chen, Y Luo, J Yi, J Bailey

Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision (ICCV) | IEEE | Published : 2020


Training accurate deep neural networks (DNNs) in the presence of noisy labels is an important and challenging 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 (CE) exhibits overfitting to noisy labels on some classes (''easy' classes), but more surprisingly, it also suffers from significant under learning on some other classes (''hard' classes). Intuitively, CE requires an extra term to facilitate learning of hard classes, and more importantly, this term should be noise tolerant, so as to avoid overfitting to noisy labels. Inspired by the symmetric KL-divergence, we prop..

View full abstract