Learning Noise Transition Matrix from Only Noisy Labels via Total Variation Regularization
This work addresses the problem of accurately estimating noise transition matrices for weakly supervised classification, which is crucial for improving model robustness and performance when only noisy labels are available.
This paper proposes a method to estimate the noise transition matrix and learn a classifier simultaneously from only noisy labels, without relying on unreliable noisy class-posterior estimation. The method introduces total variation regularization to encourage distinguishable predicted probabilities, leading to a consistent estimator of the transition matrix.
Many weakly supervised classification methods employ a noise transition matrix to capture the class-conditional label corruption. To estimate the transition matrix from noisy data, existing methods often need to estimate the noisy class-posterior, which could be unreliable due to the overconfidence of neural networks. In this work, we propose a theoretically grounded method that can estimate the noise transition matrix and learn a classifier simultaneously, without relying on the error-prone noisy class-posterior estimation. Concretely, inspired by the characteristics of the stochastic label corruption process, we propose total variation regularization, which encourages the predicted probabilities to be more distinguishable from each other. Under mild assumptions, the proposed method yields a consistent estimator of the transition matrix. We show the effectiveness of the proposed method through experiments on benchmark and real-world datasets.