FixBi: Bridging Domain Spaces for Unsupervised Domain Adaptation
arXiv:2011.09230
Abstract
Unsupervised domain adaptation (UDA) methods for learning domain invariant representations have achieved remarkable progress. However, most of the studies were based on direct adaptation from the source domain to the target domain and have suffered from large domain discrepancies. In this paper, we propose a UDA method that effectively handles such large domain discrepancies. We introduce a fixed ratio-based mixup to augment multiple intermediate domains between the source and target domain. From the augmented-domains, we train the source-dominant model and the target-dominant model that have complementary characteristics. Using our confidence-based learning methodologies, e.g., bidirectional matching with high-confidence predictions and self-penalization using low-confidence predictions, the models can learn from each other or from its own results. Through our proposed methods, the models gradually transfer domain knowledge from the source to the target domain. Extensive experiments demonstrate the superiority of our proposed method on three public benchmarks: Office-31, Office-Home, and VisDA-2017.
arXiv commentAccepted to CVPR 2021
References in corpus (9)
- Learning Transferable Features with Deep Adaptation Networks
- FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
- Conditional Adversarial Domain Adaptation
- CyCADA: Cycle-Consistent Adversarial Domain Adaptation
- VisDA: The Visual Domain Adaptation Challenge
- Model Adaptation: Unsupervised Domain Adaptation without Source Data
- ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring
- Virtual Mixup Training for Unsupervised Domain Adaptation
- Online Meta-Learning for Multi-Source and Semi-Supervised Domain Adaptation