Tripartite: Tackle Noisy Labels by a More Precise Partition
This addresses noisy label handling in deep learning, which is crucial for improving model robustness in real-world applications, though it is an incremental advancement over existing partitioning methods.
The paper tackles the problem of noisy labels in large-scale datasets by proposing a Tripartite method that partitions data into hard, noisy, and clean subsets, achieving improved performance over state-of-the-art methods on five benchmark datasets, particularly on real-world datasets.
Samples in large-scale datasets may be mislabeled due to various reasons, and Deep Neural Networks can easily over-fit to the noisy label data. To tackle this problem, the key point is to alleviate the harm of these noisy labels. Many existing methods try to divide training data into clean and noisy subsets in terms of loss values, and then process the noisy label data varied. One of the reasons hindering a better performance is the hard samples. As hard samples always have relatively large losses whether their labels are clean or noisy, these methods could not divide them precisely. Instead, we propose a Tripartite solution to partition training data more precisely into three subsets: hard, noisy, and clean. The partition criteria are based on the inconsistent predictions of two networks, and the inconsistency between the prediction of a network and the given label. To minimize the harm of noisy labels but maximize the value of noisy label data, we apply a low-weight learning on hard data and a self-supervised learning on noisy label data without using the given labels. Extensive experiments demonstrate that Tripartite can filter out noisy label data more precisely, and outperforms most state-of-the-art methods on five benchmark datasets, especially on real-world datasets.