Unsupervised Domain Adaptation via Discriminative Manifold Propagation

TL;DR

This paper introduces a Riemannian manifold learning framework for unsupervised domain adaptation that uses soft labels and global structure approximation to improve transferability and discriminability of features.

Contribution

It proposes a novel discriminative manifold propagation method that addresses pseudo label errors and global structure modeling in unsupervised domain adaptation.

Findings

Outperforms existing methods in various domain adaptation tasks.

Provides theoretical error bounds for manifold alignment metrics.

Demonstrates effectiveness in both vanilla and partial domain adaptation settings.

Abstract

Unsupervised domain adaptation is effective in leveraging rich information from a labeled source domain to an unlabeled target domain. Though deep learning and adversarial strategy made a significant breakthrough in the adaptability of features, there are two issues to be further studied. First, hard-assigned pseudo labels on the target domain are arbitrary and error-prone, and direct application of them may destroy the intrinsic data structure. Second, batch-wise training of deep learning limits the characterization of the global structure. In this paper, a Riemannian manifold learning framework is proposed to achieve transferability and discriminability simultaneously. For the first issue, this framework establishes a probabilistic discriminant criterion on the target domain via soft labels. Based on pre-built prototypes, this criterion is extended to a global approximation scheme for the second issue. Manifold metric alignment is adopted to be compatible with the embedding space. The theoretical error bounds of different alignment metrics are derived for constructive guidance. The proposed method can be used to tackle a series of variants of domain adaptation problems, including both vanilla and partial settings. Extensive experiments have been conducted to investigate the method and a comparative study shows the superiority of the discriminative manifold learning framework.