Are Anchor Points Really Indispensable in Label-Noise Learning?
This addresses a key limitation in label-noise learning for building consistent classifiers when anchor points are unavailable, offering a practical improvement for applications with noisy labels.
The paper tackles the problem of learning noise transition matrices in label-noise learning without relying on anchor points, proposing a transition-revision method that initializes the matrix using data similar to anchor points and refines it with a slack variable, achieving superior performance over state-of-the-art methods on benchmark datasets.
In label-noise learning, \textit{noise transition matrix}, denoting the probabilities that clean labels flip into noisy labels, plays a central role in building \textit{statistically consistent classifiers}. Existing theories have shown that the transition matrix can be learned by exploiting \textit{anchor points} (i.e., data points that belong to a specific class almost surely). However, when there are no anchor points, the transition matrix will be poorly learned, and those current consistent classifiers will significantly degenerate. In this paper, without employing anchor points, we propose a \textit{transition-revision} ($T$-Revision) method to effectively learn transition matrices, leading to better classifiers. Specifically, to learn a transition matrix, we first initialize it by exploiting data points that are similar to anchor points, having high \textit{noisy class posterior probabilities}. Then, we modify the initialized matrix by adding a \textit{slack variable}, which can be learned and validated together with the classifier by using noisy data. Empirical results on benchmark-simulated and real-world label-noise datasets demonstrate that without using exact anchor points, the proposed method is superior to the state-of-the-art label-noise learning methods.