DuaDiff: Dual-Conditional Diffusion Model for Guided Thermal Image Super-Resolution
- 11 December 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 37 (6) , 2809-2823
Abstract
Thermal imaging offers valuable properties, but suffers from inherently low spatial resolution, which can be enhanced using a high-resolution (HR) visible image as guidance. However, the substantial modality differences between thermal and visible images, coupled with significant resolution gaps, pose challenges to existing guided super-resolution (SR) approaches. In this article, we present dual-conditional diffusion (DuaDiff), an innovative diffusion model featuring a dual-conditioning mechanism to enhance guided thermal image SR. Unlike typical conditional diffusion models, DuaDiff integrates a learnable Laplacian pyramid to extract high-frequency details from the visible image, serving as one of the conditioning inputs. By capturing multiscale high-frequency components, DuaDiff effectively focuses on intricate textures and edges in the HR visible images, significantly enhancing thermal image fidelity. Furthermore, we project both thermal and visible images into a semantic latent space, constructing another conditioning input. Leveraging these complementary conditions, DuaDiff employs a multimodal latent feature cross-attention module to facilitate effective interaction between noise, thermal, and visible latent representations. Extensive experiments on the FLIR-ADAS and CATS datasets for $4\times $ and $8\times $ guided SR demonstrate that combining learnable Laplacian conditioning with semantic latent conditioning enables DuaDiff to surpass state-of-the-art methods in both visual quality and metric evaluation, particularly in scenarios with a large resolution gap. Besides, the applications to downstream tasks further confirm the capability of DuaDiff to recover high-fidelity semantic information. The code will be released.Keywords
Funding Information
- Natural Science Foundation of Liaoning Province (2024-MSBA-42)
- Science and Technology Major Project of Liaoning Province (2024JH1/11700048)
- Research Program of the Liaoning Liaohe Laboratory (LLL23ZZ-05-01)
- Key Research and Development Program of Liaoning Province (2023JH26/10200011)
- State Key Laboratory of Synthetical Automation for Process Industries (SAPI-2025- KFKT-02)
This publication has 56 references indexed in Scilit:
- DiffDD: A surface defect detection framework with diffusion probabilistic modelAdvanced Engineering Informatics, 2024
- Multi-modal data cross-domain fusion network for gearbox fault diagnosis under variable operating conditionsEngineering Applications of Artificial Intelligence, 2024
- Enhancement of guided thermal image super-resolution approachesNeurocomputing, 2023
- Image super-resolution: A comprehensive review, recent trends, challenges and applicationsInformation Fusion, 2022
- Super-resolution reconstruction of infrared images based on a convolutional neural network with skip connectionsOptics and Lasers in Engineering, 2021
- Pyramidal Edge-Maps and Attention Based Guided Thermal Super-Resolution2020
- Image Super-Resolution Using Very Deep Residual Channel Attention Networks2018
- Accelerating the Super-Resolution Convolutional Neural Network2016
- Learning a Deep Convolutional Network for Image Super-Resolution2014
- A universal image quality indexIEEE Signal Processing Letters, 2002