Masking: A New Perspective Of Noisy Supervision

Authors:
Bo Han RIKEN & UTS
Jiangchao Yao Shanghai Jiao Tong University
Gang Niu RIKEN
Mingyuan Zhou University of Texas at Austin
Ivor Tsang University of Technology, Sydney
Ya Zhang Cooperative Medianet Innovation Center, Shang hai Jiao Tong University
Masashi Sugiyama RIKEN / University of Tokyo

Introduction:

It is important to learn various types of classifiers given training data with noisy labels.In this paper, the authors propose a human-assisted approach called 'Masking' that conveys human cognition of invalid class transitions and naturally speculates the structure of the noise transition matrix.

Abstract:

It is important to learn various types of classifiers given training data with noisy labels. Noisy labels, in the most popular noise model hitherto, are corrupted from ground-truth labels by an unknown noise transition matrix. Thus, by estimating this matrix, classifiers can escape from overfitting those noisy labels. However, such estimation is practically difficult, due to either the indirect nature of two-step approaches, or not big enough data to afford end-to-end approaches. In this paper, we propose a human-assisted approach called ''Masking'' that conveys human cognition of invalid class transitions and naturally speculates the structure of the noise transition matrix. To this end, we derive a structure-aware probabilistic model incorporating a structure prior, and solve the challenges from structure extraction and structure alignment. Thanks to Masking, we only estimate unmasked noise transition probabilities and the burden of estimation is tremendously reduced. We conduct extensive experiments on CIFAR-10 and CIFAR-100 with three noise structures as well as the industrial-level Clothing1M with agnostic noise structure, and the results show that Masking can improve the robustness of classifiers significantly.

You may want to know: