From person to group re-identification via unsupervised transfer of sparse features
作者：
Highlights：
• We introduce an unsupervised method to tackle the group Re-ID problem.
• Dictionary learning is used to learn a representation in the single Re-ID domain.
• The learned representation is transferred to the group domain in an unsupervised way.
• A spatially invariant residual representation is used to describe a whole group image.
• Two new additional datasets have been collected to validate the proposed approach.
摘要
•We introduce an unsupervised method to tackle the group Re-ID problem.•Dictionary learning is used to learn a representation in the single Re-ID domain.•The learned representation is transferred to the group domain in an unsupervised way.•A spatially invariant residual representation is used to describe a whole group image.•Two new additional datasets have been collected to validate the proposed approach.
论文关键词：Group Re-Identification,Dictionary learning,Encoding
论文评审过程：Received 2 November 2018, Accepted 27 February 2019, Available online 16 March 2019, Version of Record 1 April 2019.
论文官网地址：https://doi.org/10.1016/j.imavis.2019.02.009
