Coupled adversarial learning for semi-supervised heterogeneous face recognition

作者：

Highlights：

• An adversarial network is proposed to generate compelling perceptual NIRVIS images.

• Two novel loss functions are designed to preserve global and local NIRVIS details.

• An end-to-end CNN network structure is proposed to learn shared features.

• The proposed semi-supervised method improves the state-of-the-art performance of HFR.

摘要

•An adversarial network is proposed to generate compelling perceptual NIRVIS images.•Two novel loss functions are designed to preserve global and local NIRVIS details.•An end-to-end CNN network structure is proposed to learn shared features.•The proposed semi-supervised method improves the state-of-the-art performance of HFR.

论文关键词：Adversarial learning,Heterogeneous face recognition,Deep representation

论文评审过程：Received 4 October 2019, Revised 8 July 2020, Accepted 24 August 2020, Available online 29 August 2020, Version of Record 1 November 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107618