Generating Thermal Human Faces for Physiological Assessment Using Thermal Sensor Auxiliary Labels

TL;DR

This paper introduces favtGAN, a novel visible-to-thermal face translation model that leverages auxiliary sensor labels and combined datasets to improve thermal face generation for medical and telemedicine applications.

Contribution

The paper presents favtGAN, a new GAN-based approach that incorporates auxiliary sensor labels and dataset fusion to enhance thermal face synthesis from visible images.

Findings

favtGAN outperforms single-dataset models in SSIM and PSNR scores

Combining face and cityscapes datasets improves thermal face generation

Auxiliary sensor label prediction enhances translation quality

Abstract

Thermal images reveal medically important physiological information about human stress, signs of inflammation, and emotional mood that cannot be seen on visible images. Providing a method to generate thermal faces from visible images would be highly valuable for the telemedicine community in order to show this medical information. To the best of our knowledge, there are limited works on visible-to-thermal (VT) face translation, and many current works go the opposite direction to generate visible faces from thermal surveillance images (TV) for law enforcement applications. As a result, we introduce favtGAN, a VT GAN which uses the pix2pix image translation model with an auxiliary sensor label prediction network for generating thermal faces from visible images. Since most TV methods are trained on only one data source drawn from one thermal sensor, we combine datasets from faces and cityscapes. These combined data are captured from similar sensors in order to bootstrap the training and transfer learning task, especially valuable because visible-thermal face datasets are limited. Experiments on these combined datasets show that favtGAN demonstrates an increase in SSIM and PSNR scores of generated thermal faces, compared to training on a single face dataset alone.