Unsupervised Domain Adaptation via Discriminative Manifold Propagation

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.

arXiv commentTo be published in IEEE Transactions on Pattern Analysis and Machine Intelligence

References in corpus (7)

• How transferable are features in deep neural networks?
• Conditional Adversarial Domain Adaptation
• VisDA: The Visual Domain Adaptation Challenge
• Dual Adversarial Auto-Encoders for Clustering
• Learning Kernel for Conditional Moment-Matching Discrepancy-based Image Classification
• Multi-step Online Unsupervised Domain Adaptation
• Discriminative Residual Analysis for Image Set Classification with Posture and Age Variations

Cited by in corpus (12)

• Multi-Source Unsupervised Domain Adaptation via Pseudo Target Domain
• Cross-Site Severity Assessment of COVID-19 from CT Images via Domain Adaptation
• HierMUD: Hierarchical Multi-task Unsupervised Domain Adaptation between Bridges for Drive-by Damage Diagnosis
• Discriminative Radial Domain Adaptation
• Towards Unsupervised Domain Adaptation via Domain-Transformer
• Sensor-invariant Fingerprint ROI Segmentation Using Recurrent Adversarial Learning
• Geometric Understanding of Discriminability and Transferability for Visual Domain Adaptation
• When Invariant Representation Learning Meets Label Shift: Insufficiency and Theoretical Insights
• Unsupervised Domain Adaptation for Image Classification via Structure-Conditioned Adversarial Learning
• Conditional Bures Metric for Domain Adaptation
• Deep Least Squares Alignment for Unsupervised Domain Adaptation
• Deep Spherical Manifold Gaussian Kernel for Unsupervised Domain Adaptation