Joint Semantic Domain Alignment and Target Classifier Learning for\n Unsupervised Domain Adaptation

2019/06/10 by Dongdong Chen, Chen, Dong-Dong, Yisen Wang +7
Computer Science · #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and ELM #Multimodal Machine Learning Applications

paper · pdf · doi:10.48550/arxiv.1906.04053

openalex publication_date 2019/06/10 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28

Abstract

Unsupervised domain adaptation aims to transfer the classifier learned from\nthe source domain to the target domain in an unsupervised manner. With the help\nof target pseudo-labels, aligning class-level distributions and learning the\nclassifier in the target domain are two widely used objectives. Existing\nmethods often separately optimize these two individual objectives, which makes\nthem suffer from the neglect of the other. However, optimizing these two\naspects together is not trivial. To alleviate the above issues, we propose a\nnovel method that jointly optimizes semantic domain alignment and target\nclassifier learning in a holistic way. The joint optimization mechanism can not\nonly eliminate their weaknesses but also complement their strengths. The\ntheoretical analysis also verifies the favor of the joint optimization\nmechanism. Extensive experiments on benchmark datasets show that the proposed\nmethod yields the best performance in comparison with the state-of-the-art\nunsupervised domain adaptation methods.\n

Citations

• Deep Domain Confusion: Maximizing for Domain Invariance
• Domain Adaptation: Learning Bounds and Algorithms
• Gradient-based learning applied to document recognition
• ImageNet Large Scale Visual Recognition Challenge
• Learning Transferable Features with Deep Adaptation Networks
• ImageNet classification with deep convolutional neural networks
• Unsupervised Domain Adaptation with Residual Transfer Networks
• Domain Separation Networks
• Domain Adaptation for Visual Applications: A Comprehensive Survey
• Asymmetric Tri-training for Unsupervised Domain Adaptation
• VisDA: The Visual Domain Adaptation Challenge
• Conditional Adversarial Domain Adaptation
• Co-regularized Alignment for Unsupervised Domain Adaptation
• Progressive Feature Alignment for Unsupervised Domain Adaptation
• Domain-Invariant Adversarial Learning for Unsupervised Domain Adaption
• Transferrable Prototypical Networks for Unsupervised Domain Adaptation

Related

• Boosting Unsupervised Domain Adaptation with Soft Pseudo-label and Curriculum Learning
• Crucial Semantic Classifier-based Adversarial Learning for Unsupervised Domain Adaptation
• Surprisingly Simple Semi-Supervised Domain Adaptation with Pretraining and Consistency
• Implicit Class-Conditioned Domain Alignment for Unsupervised Domain Adaptation
• Your Classifier can Secretly Suffice Multi-Source Domain Adaptation