Implicit Class-Conditioned Domain Alignment for Unsupervised Domain Adaptation

Abstract

We present an approach for unsupervised domain adaptation---with a strong focus on practical considerations of within-domain class imbalance and between-domain class distribution shift---from a class-conditioned domain alignment perspective. Current methods for class-conditioned domain alignment aim to explicitly minimize a loss function based on pseudo-label estimations of the target domain. However, these methods suffer from pseudo-label bias in the form of error accumulation. We propose a method that removes the need for explicit optimization of model parameters from pseudo-labels directly. Instead, we present a sampling-based implicit alignment approach, where the sample selection procedure is implicitly guided by the pseudo-labels. Theoretical analysis reveals the existence of a domain-discriminator shortcut in misaligned classes, which is addressed by the proposed implicit alignment approach to facilitate domain-adversarial learning. Empirical results and ablation studies confirm the effectiveness of the proposed approach, especially in the presence of within-domain class imbalance and between-domain class distribution shift.

Accepted at ICML2020. For code, see https://github.com/xiangdal/implicit_alignment

Cited by in corpus (12)

• CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation
• Cycle Self-Training for Domain Adaptation
• Relation Matters: Foreground-aware Graph-based Relational Reasoning for Domain Adaptive Object Detection
• Prototypical Cross-domain Self-supervised Learning for Few-shot Unsupervised Domain Adaptation
• ConDA: Continual Unsupervised Domain Adaptation
• Nearest Neighborhood-Based Deep Clustering for Source Data-absent Unsupervised Domain Adaptation
• I3Net: Implicit Instance-Invariant Network for Adapting One-Stage Object Detectors
• Prototypical Cross-domain Knowledge Transfer for Cervical Dysplasia Visual Inspection
• Multi-source Few-shot Domain Adaptation
• Omni-GAN: On the Secrets of cGANs and Beyond
• Adversarial Robustness for Unsupervised Domain Adaptation
• Adaptive Pseudo-Label Refinement by Negative Ensemble Learning for Source-Free Unsupervised Domain Adaptation