Multi-level feature fusion and multi-loss learning for person Re-Identification

作者：

Highlights：

• The self-attention model is used to weight the importance of different regions.

• The multi-loss function is used to enhance the feature learning ability.

• Our method achieves better experimental results on three challenging datasets

摘要

•The self-attention model is used to weight the importance of different regions.•The multi-loss function is used to enhance the feature learning ability.•Our method achieves better experimental results on three challenging datasets

论文关键词：Self-attention module,Relative weight,Multi-loss learning

论文评审过程：Received 24 March 2020, Revised 9 January 2021, Accepted 7 February 2021, Available online 22 February 2021, Version of Record 26 February 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116197