GearNet: Stepwise Dual Learning for Weakly Supervised Domain Adaptation

TL;DR

GearNet introduces a bilateral learning framework for weakly supervised domain adaptation, effectively leveraging mutual domain relationships to improve performance despite noisy labels in the source domain.

Contribution

It proposes a novel bilateral learning paradigm with asymmetrical KL loss, enhancing existing WSDA methods by exploiting domain correlations.

Findings

Significant performance improvements on benchmark datasets.

Effective implicit noise cancellation in source labels.

Versatile enhancement applicable to various WSDA methods.

Abstract

This paper studies weakly supervised domain adaptation(WSDA) problem, where we only have access to the source domain with noisy labels, from which we need to transfer useful information to the unlabeled target domain. Although there have been a few studies on this problem, most of them only exploit unidirectional relationships from the source domain to the target domain. In this paper, we propose a universal paradigm called GearNet to exploit bilateral relationships between the two domains. Specifically, we take the two domains as different inputs to train two models alternately, and asymmetrical Kullback-Leibler loss is used for selectively matching the predictions of the two models in the same domain. This interactive learning schema enables implicit label noise canceling and exploits correlations between the source and target domains. Therefore, our GearNet has the great potential to boost the performance of a wide range of existing WSDL methods. Comprehensive experimental results show that the performance of existing methods can be significantly improved by equipping with our GearNet.