Robust Learning by Self-Transition for Handling Noisy Labels
This work provides a method to improve the robustness and efficiency of deep neural networks for practitioners dealing with real-world datasets containing noisy labels.
This paper addresses the problem of noisy labels in real-world data, which degrades the generalization of deep neural networks. The proposed method, MORPH, automatically switches its learning phase from seeding to evolution, effectively avoiding overfitting to false-labeled samples. Experiments on five benchmark datasets show substantial improvements over state-of-the-art methods.
Real-world data inevitably contains noisy labels, which induce the poor generalization of deep neural networks. It is known that the network typically begins to rapidly memorize false-labeled samples after a certain point of training. Thus, to counter the label noise challenge, we propose a novel self-transitional learning method called MORPH, which automatically switches its learning phase at the transition point from seeding to evolution. In the seeding phase, the network is updated using all the samples to collect a seed of clean samples. Then, in the evolution phase, the network is updated using only the set of arguably clean samples, which precisely keeps expanding by the updated network. Thus, MORPH effectively avoids the overfitting to false-labeled samples throughout the entire training period. Extensive experiments using five real-world or synthetic benchmark datasets demonstrate substantial improvements over state-of-the-art methods in terms of robustness and efficiency.