Adversarial Unsupervised Domain Adaptation with Conditional and Label Shift: Infer, Align and Iterate
This addresses domain adaptation challenges for machine learning applications where label distributions shift across domains, offering a practical solution but with incremental improvements over existing methods.
The paper tackles unsupervised domain adaptation under conditional and label shifts by proposing an adversarial approach that iteratively infers and aligns distributions, achieving state-of-the-art results on classification and segmentation tasks with concrete performance gains.
In this work, we propose an adversarial unsupervised domain adaptation (UDA) approach with the inherent conditional and label shifts, in which we aim to align the distributions w.r.t. both $p(x|y)$ and $p(y)$. Since the label is inaccessible in the target domain, the conventional adversarial UDA assumes $p(y)$ is invariant across domains, and relies on aligning $p(x)$ as an alternative to the $p(x|y)$ alignment. To address this, we provide a thorough theoretical and empirical analysis of the conventional adversarial UDA methods under both conditional and label shifts, and propose a novel and practical alternative optimization scheme for adversarial UDA. Specifically, we infer the marginal $p(y)$ and align $p(x|y)$ iteratively in the training, and precisely align the posterior $p(y|x)$ in testing. Our experimental results demonstrate its effectiveness on both classification and segmentation UDA, and partial UDA.