Covariance descriptor based on bio-inspired features for person re-identification and face verification

作者：

Highlights：

• This paper proposes a novel person/face image representation.

• The representation avoids the use of body segmentation or image normalization.

• The representation relies on the combination of BIF and Covariance descriptor.

• The representation can handle background and illumination variations.

• The matching rate at rank 1 on VIPeR is 31.11% and the accuracy on LFW is 84.48%.

摘要

•This paper proposes a novel person/face image representation.•The representation avoids the use of body segmentation or image normalization.•The representation relies on the combination of BIF and Covariance descriptor.•The representation can handle background and illumination variations.•The matching rate at rank 1 on VIPeR is 31.11% and the accuracy on LFW is 84.48%.

论文关键词：Image representation,Person re-identification,Face verification,Biologically inspired features,Covariance descriptor

论文评审过程：Received 9 August 2013, Revised 10 January 2014, Accepted 2 April 2014, Available online 12 April 2014.

论文官网地址：https://doi.org/10.1016/j.imavis.2014.04.002