Heterogeneous Face Recognition via Face Synthesis With Identity-Attribute Disentanglement

Abstract

Heterogeneous Face Recognition (HFR) aims to match faces across different domains (e.g., visible to near-infrared images), which has been widely applied in authentication and forensics scenarios. However, HFR is a challenging problem because of the large cross-domain discrepancy, limited heterogeneous data pairs, and large variation of facial attributes.To address these challenges, we propose a new HFR method from the perspective of heterogeneous data augmentation, named Face Synthesis with Identity-Attribute Disentanglement (FSIAD). Firstly, the identity-attribute disentanglement (IAD) decouples face images into identity-related representations and identityunrelated representations (called attributes), and then decreases the correlation between identities and attributes. Secondly, we devise a face synthesis module (FSM) to generate a large number of images with stochastic combinations of disentangled identities and attributes for enriching the attribute diversity of synthetic images. Both the original images and the synthetic ones are utilized to train the HFR network for tackling the challenges and improving the performance of HFR. Extensive experiments on five HFR databases validate that FSIAD obtains superior performance than previous HFR approaches. Particularly, FSIAD obtains 4.8% improvement over state of the art in terms of VR@FAR=0.01% on LAMP-HQ, the largest HFR database so far.

Unpaired Caricature-Visual Face Recognition via Feature Decomposition-Restoration-Decomposition

“…Modality-specific information compensation-based methods [1], [4], [18] try to compensate for the missing modalityspecific information in each modality. Yang et al [4] use a generative adversarial network to generate facial images, enriching the attribute diversity of synthetic images. DVG-Face [1] generates heterogeneous facial images with the same identity from noise.…”

“…O VER the past few years, cross-modality face recognition has received significant attention due to the rapid growth of multi-modality data. Accordingly, a number of methods [1]- [4] have been developed and achieved promising performance. These efforts stem from the growing demand for advanced face recognition technologies that can operate across diverse Fig.…”