This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Open AccessArticle
A Dual-Branch Lightweight Network for Multimodal Image Fusion with Mamba and INN
by
Nan Li
Nan Li
Nan Li is a master’s student majoring in Radio Physics at the School of Electronic Engineering, in [...]
Nan Li is a master’s student majoring in Radio Physics at the School of Electronic Engineering, Ili Normal University, China. His research interests include multimodal image fusion, sensor image processing, wireless signal analysis, and related applications in electronic engineering. During his postgraduate studies, he has been engaged in research on image fusion methods and sensing information processing. He is currently focusing on multimodal image fusion and its applications in sensor-based imaging systems, with the aim of improving image quality, information integration, and practical performance in complex sensing environments.
,
Hongxin Li
Hongxin Li
Hongxin Li is a master’s student majoring in Radio Physics at the School of Electronic Ili Normal [...]
Hongxin Li is a master’s student majoring in Radio Physics at the School of Electronic Engineering, Ili Normal University, China. He is the second author of a manuscript submitted to Sensors. His academic background and current research are related to electronic engineering, radio physics, sensor information processing, and infrared small target detection. During his postgraduate studies, he has participated in research work involving infrared image analysis, target detection methods, and signal processing applications. He is currently focusing on improving the accuracy and robustness of infrared small target detection in complex sensing environments.
Lin Tian
Lin Tian
Lin Tian is an Associate Professor and master’s supervisor in Radio Physics at the School of Ili [...]
Lin Tian is an Associate Professor and master’s supervisor in Radio Physics at the School of Electronic Engineering, Ili Normal University, China. She is pursuing her doctoral degree at the University of Electronic Science and Technology of China. She serves as the corresponding author of a manuscript submitted to Sensors. Her research interests include multimodal image fusion, infrared small target detection, seismic data reconstruction, sensor information processing, and related applications in electronic engineering. She has been engaged in teaching and research in radio physics and electronic engineering, with a focus on intelligent sensing, image analysis, and data reconstruction methods for complex sensing environments.
1
Xinjiang Laboratory of Phase Transitions and Microstructures in Condensed Matter Physics, Yili Normal University, Yining 835000, China
2
School of Electronic Engineering, Yili Normal University, Yining 835000, China
*
Author to whom correspondence should be addressed.
Sensors 2026, 26(12), 3814; https://doi.org/10.3390/s26123814 (registering DOI)
Submission received: 11 May 2026
/
Revised: 7 June 2026
/
Accepted: 12 June 2026
/
Published: 15 June 2026
Abstract
Multimodal image fusion aims to integrate complementary information from heterogeneous imaging modalities into a single informative image. However, many deep learning-based fusion methods rely on complex feature extractors, leading to high computational cost and limited suitability for real-time deployment on resource-constrained devices. To address this issue, this paper proposes a lightweight Mamba-INN dual-branch network for efficient multimodal image fusion. The proposed model decouples global structure modeling from local detail preservation. A simplified Mamba-inspired branch is designed to capture long-range contextual dependencies, while a lightweight invertible neural network branch preserves high-frequency textures and edge information through information-preserving transformations. The lightweight INN branch preserves high-frequency texture and edge information during the forward feature transformation process through reversible feature partitioning, coupled transformations, and exponential scale modulation, thereby reducing the loss of detail caused by feature compression. Compact shallow feature refinement, module reuse, low-dimensional channel design, and a streamlined decoder are further introduced to reduce redundant computation. Experiments on infrared-visible and medical image fusion benchmarks, including MSRS, TNO, RoadScene, MRI-CT, MRI-PET, and MRI-SPECT datasets, demonstrate that the proposed method achieves competitive fusion quality with low model complexity. The proposed method achieves performance comparable to or better than that of methods such as CDDFuse, U2Fusion, CNN and SDNet on metrics including MI, VIF, Qabf, and SSIM for infrared-visible and medical image fusion tasks, while containing only 0.24 million parameters and requiring 24.04 GFLOPs of computational power at an input resolution of 256 × 256. Compared to CDDFuse, our method significantly reduces model complexity, enhancing the potential for lightweight deployment while maintaining fusion quality.
Share and Cite
MDPI and ACS Style
Li, N.; Li, H.; Tian, L.
A Dual-Branch Lightweight Network for Multimodal Image Fusion with Mamba and INN. Sensors 2026, 26, 3814.
https://doi.org/10.3390/s26123814
AMA Style
Li N, Li H, Tian L.
A Dual-Branch Lightweight Network for Multimodal Image Fusion with Mamba and INN. Sensors. 2026; 26(12):3814.
https://doi.org/10.3390/s26123814
Chicago/Turabian Style
Li, Nan, Hongxin Li, and Lin Tian.
2026. "A Dual-Branch Lightweight Network for Multimodal Image Fusion with Mamba and INN" Sensors 26, no. 12: 3814.
https://doi.org/10.3390/s26123814
APA Style
Li, N., Li, H., & Tian, L.
(2026). A Dual-Branch Lightweight Network for Multimodal Image Fusion with Mamba and INN. Sensors, 26(12), 3814.
https://doi.org/10.3390/s26123814
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details
here.
Article Metrics
Article metric data becomes available approximately 24 hours after publication online.
