Adversarial Bi-Regressor Network for Domain Adaptive Regression

TL;DR

This paper introduces ABRNet, a novel adversarial bi-regressor approach for domain adaptive regression, effectively reducing domain shift and improving cross-domain regression tasks like indoor localization.

Contribution

The paper proposes a new bi-regressor architecture combined with adversarial training and domain-specific augmentation to enhance domain adaptation in regression tasks.

Findings

Effective in reducing domain gap in regression tasks

Outperforms existing methods on benchmark datasets

Improves accuracy in indoor localization applications

Abstract

Domain adaptation (DA) aims to transfer the knowledge of a well-labeled source domain to facilitate unlabeled target learning. When turning to specific tasks such as indoor (Wi-Fi) localization, it is essential to learn a cross-domain regressor to mitigate the domain shift. This paper proposes a novel method Adversarial Bi-Regressor Network (ABRNet) to seek more effective cross-domain regression model. Specifically, a discrepant bi-regressor architecture is developed to maximize the difference of bi-regressor to discover uncertain target instances far from the source distribution, and then an adversarial training mechanism is adopted between feature extractor and dual regressors to produce domain-invariant representations. To further bridge the large domain gap, a domain-specific augmentation module is designed to synthesize two source-similar and target-similar intermediate domains to gradually eliminate the original domain mismatch. The empirical studies on two cross-domain regressive benchmarks illustrate the power of our method on solving the domain adaptive regression (DAR) problem.