Heterogeneous Face Frontalization via Domain Agnostic Learning

TL;DR

This paper introduces DAL-GAN, a novel domain agnostic GAN model that effectively synthesizes frontal visible faces from thermal images with pose variations, improving biometric and surveillance applications.

Contribution

The paper proposes a new DAL-GAN framework with dual discriminators, contrastive latent space constraints, and a multi-loss function for improved thermal to visible face frontalization.

Findings

DAL-GAN outperforms baseline methods in quality of synthesized faces.

The model preserves identity across modalities and poses.

Extensive experiments validate the effectiveness of the proposed approach.

Abstract

Recent advances in deep convolutional neural networks (DCNNs) have shown impressive performance improvements on thermal to visible face synthesis and matching problems. However, current DCNN-based synthesis models do not perform well on thermal faces with large pose variations. In order to deal with this problem, heterogeneous face frontalization methods are needed in which a model takes a thermal profile face image and generates a frontal visible face. This is an extremely difficult problem due to the large domain as well as large pose discrepancies between the two modalities. Despite its applications in biometrics and surveillance, this problem is relatively unexplored in the literature. We propose a domain agnostic learning-based generative adversarial network (DAL-GAN) which can synthesize frontal views in the visible domain from thermal faces with pose variations. DAL-GAN consists of a generator with an auxiliary classifier and two discriminators which capture both local and global texture discriminations for better synthesis. A contrastive constraint is enforced in the latent space of the generator with the help of a dual-path training strategy, which improves the feature vector discrimination. Finally, a multi-purpose loss function is utilized to guide the network in synthesizing identity preserving cross-domain frontalization. Extensive experimental results demonstrate that DAL-GAN can generate better quality frontal views compared to the other baseline methods.