Label propagation with multi-stage inference for visual domain adaptation

作者：

Highlights：

• Discriminative structure in labeled source domain and manifold structure in unlabeled target domain are embodied simultaneously.

• We propose to make selective connection for cross-domain similarity graph construction.

• We propose multi-stage inference for predicting target labels.

摘要

•Discriminative structure in labeled source domain and manifold structure in unlabeled target domain are embodied simultaneously.•We propose to make selective connection for cross-domain similarity graph construction.•We propose multi-stage inference for predicting target labels.

论文关键词：Domain adaptation,Label propagation,Classifier adaptation

论文评审过程：Received 11 August 2020, Revised 16 November 2020, Accepted 23 January 2021, Available online 29 January 2021, Version of Record 3 February 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.106809