Learning adaptive geometry for unsupervised domain adaptation
作者：
Highlights：
• Identify the problem of geometry difference in unsupervised domain adaptation.
• Propose learning adaptive geometry for geometry alignment without target labels.
• A geometry-aware dual-stream network to learn the geometry-aligned representations.
• Design unified geometry criteria as losses for adaptive geometry learning.
• Achieve good performance in several cross-dataset recognition tasks.
摘要
•Identify the problem of geometry difference in unsupervised domain adaptation.•Propose learning adaptive geometry for geometry alignment without target labels.•A geometry-aware dual-stream network to learn the geometry-aligned representations.•Design unified geometry criteria as losses for adaptive geometry learning.•Achieve good performance in several cross-dataset recognition tasks.
论文关键词：Domain adaptation,Manifold structure,Distribution alignment
论文评审过程：Received 18 March 2020, Revised 15 July 2020, Accepted 6 September 2020, Available online 14 September 2020, Version of Record 18 September 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107638
