TherA：热感知视觉语言提示用于可控 RGB 到热红外翻译
TherA: Thermal-Aware Visual-Language Prompting for Controllable RGB-to-Thermal Infrared Translation
TherA 将视觉语言模型（VLM）作为条件编码器，把用户提示的场景-物体条件对映射为热感知嵌入，再驱动潜扩散模型实现可控的 RGB 到热红外翻译；这种“VLM 编码条件 + 扩散解码”的范式可复用到其他需要物理属性感知的跨模态图像翻译任务（如 RGB 到近红外、深度、事件相机）。
原文依据 TherA couples TherA-VLM with a latent-diffusion-based translator.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论尽管热红外（TIR）成像具有固有优势，但大规模数据采集与标注仍是基于 TIR 感知的主要瓶颈。一种实用的替代方案是通过图像翻译合成伪 TIR 数据；然而，大多数 RGB 到 TIR 方法严重依赖以 RGB 为中心的先验，忽略了热物理特性，导致热分布不合理。本文提出 TherA，一个可控的 RGB 到 TIR 翻译框架，能够在场景和物体级别生成多样且热合理的图像。TherA 将 TherA-VLM 与基于潜扩散的翻译器耦合。给定单张 RGB 图像和用户提示的条件对，TherA-VLM 生成一个热感知嵌入，该嵌入编码了反映输入场景-条件对的场景、物体、材质和热辐射上下文。以该嵌入为条件驱动扩散模型，可实现逼真的 TIR 合成，并在一天中的时间、天气和物体状态上实现细粒度控制。与其他基线相比，TherA 取得了最优的翻译性能，在所有指标上平均零样本翻译性能提升高达 33%。
英文原文摘要 1174 字符
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.