English

Attention-based Cross-Layer Domain Alignment for Unsupervised Domain Adaptation

Computer Vision and Pattern Recognition 2022-03-01 v1

Abstract

Unsupervised domain adaptation (UDA) aims to learn transferable knowledge from a labeled source domain and adapts a trained model to an unlabeled target domain. To bridge the gap between source and target domains, one prevailing strategy is to minimize the distribution discrepancy by aligning their semantic features extracted by deep models. The existing alignment-based methods mainly focus on reducing domain divergence in the same model layer. However, the same level of semantic information could distribute across model layers due to the domain shifts. To further boost model adaptation performance, we propose a novel method called Attention-based Cross-layer Domain Alignment (ACDA), which captures the semantic relationship between the source and target domains across model layers and calibrates each level of semantic information automatically through a dynamic attention mechanism. An elaborate attention mechanism is designed to reweight each cross-layer pair based on their semantic similarity for precise domain alignment, effectively matching each level of semantic information during model adaptation. Extensive experiments on multiple benchmark datasets consistently show that the proposed method ACDA yields state-of-the-art performance.

Keywords

Cite

@article{arxiv.2202.13310,
  title  = {Attention-based Cross-Layer Domain Alignment for Unsupervised Domain Adaptation},
  author = {Xu Ma and Junkun Yuan and Yen-wei Chen and Ruofeng Tong and Lanfen Lin},
  journal= {arXiv preprint arXiv:2202.13310},
  year   = {2022}
}

Comments

Accepted by Neurocomputing

R2 v1 2026-06-24T09:55:15.035Z