Masking: A New Perspective of Noisy Supervision
This work addresses the challenge of noisy label supervision in machine learning, which is crucial for improving model reliability in real-world applications, though it is incremental as it builds on existing noise transition matrix estimation methods.
The paper tackles the problem of learning classifiers from training data with noisy labels by proposing Masking, a human-assisted approach that incorporates human cognition of invalid class transitions to estimate the noise transition matrix, resulting in significant improvements in classifier robustness on CIFAR-10, CIFAR-100, and Clothing1M datasets.
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.