Synthetic thermal data boosts visible to thermal face recognition
SynThermFace: Amplifying Limited Paired Data for Visible-Thermal Face Recognition via Synthetic Data Generation
Computer Vision and Pattern Recognition
Summary
Face recognition usually works best with normal visible-light photos but struggles in the dark or low-light conditions. This paper shows how using computer-generated thermal images paired with visible images can improve face recognition when real thermal data is scarce. The researchers adapted a model to create synthetic thermal images from visible faces, which helps train recognition systems better. This method speeds up recognition since it doesn't require image conversion during use, and it works across different face datasets.
What this means in practice
- •For security system developers: Improve biometric access control by training face recognition systems that match visible and thermal images using synthetic thermal data.
- •For surveillance technology engineers: Enhance night-time face recognition by adapting models with synthetic paired visible-thermal data to maintain accuracy without extra image processing at inference.
Authors
Anjith George, Adam Unal, Sebastien Marcel
Abstract
Face recognition (FR) is a widely used modality for biometric authentication, but conventional models rely on visible-spectrum imagery and degrade when high-quality RGB images cannot be captured. Cross-spectral face recognition addresses this limitation by matching visible images with other modalities such as thermal imagery, enabling more reliable performance in low-light, nighttime, and unconstrained conditions. However, progress is limited by the scarcity of paired visible-thermal data, which is difficult and costly to collect at scale. We propose SynThermFace, a framework that amplifies limited real visible-thermal supervision into larger paired adaptation datasets for cross-spectral face recognition. A diffusion model is first adapted using a limited set of paired visible--thermal images and then used to generate large-scale paired visible--synthetic thermal data from existing real or synthetic visible face datasets. The generated pairs are used to adapt a pretrained visible-spectrum face recognition model into a CFR model. Unlike synthesis-based approaches that require image translation at test time, the proposed method shifts generation to the training stage and performs inference with a single forward pass through the adapted recognition model. Under the same MCXFace real-pair protocol, PACT improves over the evaluated CFR adaptation baselines, isolating the effect of the proposed adaptation objective. Training PACT on larger generated paired datasets provides additional improvements over both the unadapted model and the real-pair PACT configuration. Cross-database evaluation on the Tufts dataset provides evidence that the learned representation transfers to an unseen database. The source code and trained models will be made publicly available.