2020/03/30 by Jianfei Yang, Han Zou, Yang, Jianfei +5 · 9 citations
Computer Science · Mathematics · Medicine · #Adversarial system #Algorithm #Artificial intelligence #COVID-19 diagnosis using AI #Computer science #Domain (mathematical analysis) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Geography #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine learning #Margin (machine learning) #Mathematical analysis #Mathematics #Multimodal Machine Learning Applications #Training (meteorology) #Training set #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.2003.13249
published in arXiv (Cornell University) (Cornell University)
openalex created_date 2020/02/07 · arxiv created 2020/03/30 · openalex publication_date 2020/03/30 · arxiv updated 2020/03/31 · openalex updated_date 2026/07/28
Domain adaptation tackles the problem of transferring knowledge from a label-rich source domain to a label-scarce or even unlabeled target domain. Recently domain-adversarial training (DAT) has shown promising capacity to learn a domain-invariant feature space by reversing the gradient propagation of a domain classifier. However, DAT is still vulnerable in several aspects including (1) training instability due to the overwhelming discriminative ability of the domain classifier in adversarial training, (2) restrictive feature-level alignment, and (3) lack of interpretability or systematic explanation of the learned feature space. In this paper, we propose a novel Max-margin Domain-Adversarial Training (MDAT) by designing an Adversarial Reconstruction Network (ARN). The proposed MDAT stabilizes the gradient reversing in ARN by replacing the domain classifier with a reconstruction network, and in this manner ARN conducts both feature-level and pixel-level domain alignment without involving extra network structures. Furthermore, ARN demonstrates strong robustness to a wide range of hyper-parameters settings, greatly alleviating the task of model selection. Extensive empirical results validate that our approach outperforms other state-of-the-art domain alignment methods. Moreover, reconstructing adapted features reveals the domain-invariant feature space which conforms with our intuition.