Spatial-driven features based on image dependencies for person re-identification

作者：

Highlights：

• We design a GAM to capture the inter-image dependencies among a series of different pedestrian images.

• We present a LAM to compute the intra-image dependencies from any pair of pixels within each pedestrian image.

• We propose a specific network integration mechanism to match well the solution of the spatial dependency problem.

• Extensive experiments verify that the proposed method exceeds the state-of-the-art methods.

摘要

•We design a GAM to capture the inter-image dependencies among a series of different pedestrian images.•We present a LAM to compute the intra-image dependencies from any pair of pixels within each pedestrian image.•We propose a specific network integration mechanism to match well the solution of the spatial dependency problem.•Extensive experiments verify that the proposed method exceeds the state-of-the-art methods.

论文关键词：Person re-identification,Spatial dependencies,Recurrent neural network,Deep learning

论文评审过程：Received 22 March 2021, Revised 16 November 2021, Accepted 26 November 2021, Available online 27 November 2021, Version of Record 8 December 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108462