TherA: Thermal-Aware Visual-Language Prompting for Controllable RGB-to-Thermal Infrared Translation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lee, Dong-Guw, Rhee, Tai Hyoung, Jang, Hyunsoo, Shin, Young-Sik, Shin, Ukcheol, Kim, Ayoung
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914350597406720
author Lee, Dong-Guw
Rhee, Tai Hyoung
Jang, Hyunsoo
Shin, Young-Sik
Shin, Ukcheol
Kim, Ayoung
author_facet Lee, Dong-Guw
Rhee, Tai Hyoung
Jang, Hyunsoo
Shin, Young-Sik
Shin, Ukcheol
Kim, Ayoung
contents Despite the inherent advantages of thermal infrared(TIR) imaging, large-scale data collection and annotation remain a major bottleneck for TIR-based perception. A practical alternative is to synthesize pseudo TIR data via image translation; however, most RGB-to-TIR approaches heavily rely on RGB-centric priors that overlook thermal physics, yielding implausible heat distributions. In this paper, we introduce TherA, a controllable RGB-to-TIR translation framework that produces diverse and thermally plausible images at both scene and object level. TherA couples TherA-VLM with a latent-diffusion-based translator. Given a single RGB image and a user-prompted condition pair, TherA-VLM yields a thermal-aware embedding that encodes scene, object, material, and heat-emission context reflecting the input scene-condition pair. Conditioning the diffusion model on this embedding enables realistic TIR synthesis and fine-grained control across time of day, weather, and object state. Compared to other baselines, TherA achieves state-of-the-art translation performance, demonstrating improved zero-shot translation performance up to 33% increase averaged across all metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2602_19430
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle TherA: Thermal-Aware Visual-Language Prompting for Controllable RGB-to-Thermal Infrared Translation
Lee, Dong-Guw
Rhee, Tai Hyoung
Jang, Hyunsoo
Shin, Young-Sik
Shin, Ukcheol
Kim, Ayoung
Computer Vision and Pattern Recognition
Despite the inherent advantages of thermal infrared(TIR) imaging, large-scale data collection and annotation remain a major bottleneck for TIR-based perception. A practical alternative is to synthesize pseudo TIR data via image translation; however, most RGB-to-TIR approaches heavily rely on RGB-centric priors that overlook thermal physics, yielding implausible heat distributions. In this paper, we introduce TherA, a controllable RGB-to-TIR translation framework that produces diverse and thermally plausible images at both scene and object level. TherA couples TherA-VLM with a latent-diffusion-based translator. Given a single RGB image and a user-prompted condition pair, TherA-VLM yields a thermal-aware embedding that encodes scene, object, material, and heat-emission context reflecting the input scene-condition pair. Conditioning the diffusion model on this embedding enables realistic TIR synthesis and fine-grained control across time of day, weather, and object state. Compared to other baselines, TherA achieves state-of-the-art translation performance, demonstrating improved zero-shot translation performance up to 33% increase averaged across all metrics.
title TherA: Thermal-Aware Visual-Language Prompting for Controllable RGB-to-Thermal Infrared Translation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2602.19430