Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Normalized Loss Functions for Deep Learning with Noisy Labels

About

Robust loss functions are essential for training accurate deep neural networks (DNNs) in the presence of noisy (incorrect) labels. It has been shown that the commonly used Cross Entropy (CE) loss is not robust to noisy labels. Whilst new loss functions have been designed, they are only partially robust. In this paper, we theoretically show by applying a simple normalization that: any loss can be made robust to noisy labels. However, in practice, simply being robust is not sufficient for a loss function to train accurate DNNs. By investigating several robust loss functions, we find that they suffer from a problem of underfitting. To address this, we propose a framework to build robust loss functions called Active Passive Loss (APL). APL combines two robust loss functions that mutually boost each other. Experiments on benchmark datasets demonstrate that the family of new loss functions created by our APL framework can consistently outperform state-of-the-art methods by large margins, especially under large noise rates such as 60% or 80% incorrect labels.

Xingjun Ma, Hanxun Huang, Yisen Wang, Simone Romano, Sarah Erfani, James Bailey• 2020

Related benchmarks

TaskDatasetResultRank
Image ClassificationCIFAR-100 (test)
Accuracy73.88
3518
Node ClassificationCora
Accuracy76
1215
Image ClassificationImageNet (val)
Top-1 Acc63.96
1206
Node ClassificationCiteseer
Accuracy60.7
931
Image ClassificationCIFAR-10 (test)
Accuracy90.91
906
Image ClassificationMNIST (test)
Accuracy99.39
894
Node ClassificationCora (test)
Mean Accuracy72.37
861
Node ClassificationPubmed
Accuracy71.1
819
Image ClassificationClothing1M (test)
Accuracy72.18
574
Skin Lesion SegmentationISIC 2017 (test)
Dice Score82.9
113
Showing 10 of 32 rows

Other info

Follow for update