Towards Understanding Deep Learning from Noisy Labels with Small-Loss Criterion
This work addresses the challenge of noisy labels in deep learning, which is common in real-world applications like crowdsourcing, by offering a theoretical foundation and improved method, though it is incremental as it builds on existing small-loss approaches.
The paper tackles the problem of deep learning with noisy labels by providing a theoretical explanation for why the small-loss criterion works and reformulating it to improve performance, with experimental results verifying the explanation and demonstrating effectiveness.
Deep neural networks need large amounts of labeled data to achieve good performance. In real-world applications, labels are usually collected from non-experts such as crowdsourcing to save cost and thus are noisy. In the past few years, deep learning methods for dealing with noisy labels have been developed, many of which are based on the small-loss criterion. However, there are few theoretical analyses to explain why these methods could learn well from noisy labels. In this paper, we theoretically explain why the widely-used small-loss criterion works. Based on the explanation, we reformalize the vanilla small-loss criterion to better tackle noisy labels. The experimental results verify our theoretical explanation and also demonstrate the effectiveness of the reformalization.