Linfeng Tang
计算机视觉 | 图像融合 | 图像增强
我是 唐霖峰,现为武汉大学机器人学院博士后,合作导师为 马佳义教授。
我于 2025 年 12 月获武汉大学电子信息学院博士学位,于 2020 年获中南大学学士学位。我的研究方向主要包括多模态图像融合、视频融合、低照度图像增强、图像恢复,以及语义感知建模。
目前共有 6 篇 ESI 热点论文(前 0.1%)和 8 篇 ESI 高被引论文(前 1%):Mask-DiFuser、SeAFusion、PIAFusion、SwinFusion、SuperFusion、PSFusion、DIVFusion 和 STDFusionNet。
研究方向
多模图像融合
红外-可见光图像融合、视频融合、可控融合,以及配准与融合联合建模。
图像增强与恢复
低照度图像增强、退化感知恢复,以及基于扩散模型的图像恢复方法。
任务驱动感知
语义引导、语言-视觉先验,以及面向检测、分割等高层视觉任务的图像融合模型。
代表性工作
Mask-DiFuser
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI),2026
以掩码扩散模型统一建模多种无监督图像融合任务。
C2RF
International Journal of Computer Vision (IJCV),2025
ControlFusion
Advances in Neural Information Processing Systems (NeurIPS),2025
VideoFusion
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),2026
SwinFusion
IEEE/CAA Journal of Automatica Sinica,2022
利用 Swin Transformer 建模跨域长程依赖,实现通用图像融合。
SeAFusion
Information Fusion,2022
将语义感知与高层视觉任务反馈引入实时红外-可见光融合。
PIAFusion
Information Fusion,2022
依据光照条件渐进融合红外与可见光信息。
SuperFusion
IEEE/CAA Journal of Automatica Sinica,2022
联合图像配准与语义感知融合,面向空间未对齐的多模态输入。
DIVFusion
Information Fusion,2023
面向低照度场景抑制暗区并增强融合细节可见性。
PSFusion
Information Fusion,2023
通过语义注入与场景保真协同优化低层融合质量和高层任务。
STDFusionNet
IEEE Transactions on Instrumentation and Measurement,2021
利用显著目标检测保留关键热目标与结构信息。
Deep Learning-based Image Fusion: A Survey
《中国图象图形学报》,2023
本页展示的引用数据已根据本人 Google Scholar 主页于 2026 年 7 月 18 日更新,实时数据请以 Google Scholar 为准。
学术服务
目前担任 TPAMI、TIP、TNNLS、TMM、TCSVT、TII、TGRS、TITS、TSMC、TASE、JSAC、JAS、IoTJ、IJCV、Information Fusion、PR,以及 CVPR、NeurIPS、ECCV、AAAI、IJCAI、ACM MM 等期刊和会议审稿人。完整列表见 学术服务 页面。
news
| Feb 20, 2026 | Our paper “VideoFusion” accepted by CVPR 2026! [Paper][Code] |
|---|---|
| Dec 26, 2025 | 我们的综述《基于深度学习的图像融合方法综述》荣获 中国图象图形学报 2020-2024 优秀论文! |
| Sep 18, 2025 | Our paper “ControlFusion” accepted by NeurIPS 2025 (Oral)! [Paper][Code] |
| Sep 10, 2025 | Our paper “Mask-DiFuser” accepted by IEEE TPAMI! [Paper][Code] |
| Mar 15, 2025 | Our paper C2RF has been accepted by IJCV! [Paper] [Code] |
selected publications
- IEEE TPAMI
Mask-DiFuser: A Masked Diffusion Model for Unified Unsupervised Image FusionIEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2026Masked diffusion modeling that unifies multiple unsupervised image-fusion tasks; Google Scholar cited by 51 (Jul. 18, 2026); ESI Highly Cited Paper - Inf. Fusion
Image fusion in the loop of high-level vision tasks: A semantic-aware real-time infrared and visible image fusion networkInformation Fusion, 2022Semantic-aware real-time infrared-visible fusion with high-level task feedback; Information Fusion Best Paper Award 2024; Google Scholar cited by 1183 (Jul. 18, 2026); ESI Hot Paper; ESI Highly Cited Paper - IEEE/CAA JAS
SwinFusion: Cross-domain Long-range Learning for General Image Fusion via Swin TransformerIEEE/CAA Journal of Automatica Sinica, 2022Cross-domain long-range modeling with a Swin Transformer for general image fusion; IEEE/CAA JAS 2023 Best Paper Award, sole annual winner; Google Scholar cited by 1582 (Jul. 18, 2026); ESI Hot Paper; ESI Highly Cited Paper - IEEE/CAA JAS
SuperFusion: A Versatile Image Registration and Fusion Network with Semantic AwarenessIEEE/CAA Journal of Automatica Sinica, 2022Joint image registration and semantics-aware fusion for spatially misaligned inputs; Google Scholar cited by 504 (Jul. 18, 2026); ESI Hot Paper; ESI Highly Cited Paper - Inf. Fusion
Rethinking the necessity of image fusion in high-level vision tasks: A practical infrared and visible image fusion network based on progressive semantic injection and scene fidelityInformation Fusion, 2023Progressive semantic injection and scene fidelity for task-oriented fusion; Google Scholar cited by 410 (Jul. 18, 2026); ESI Hot Paper; ESI Highly Cited Paper - IEEE TIM
STDFusionNet: An Infrared and Visible Image Fusion Network Based on Salient Target DetectionIEEE Transactions on Instrumentation and Measurement, 2021Salient-target detection for preserving critical thermal targets and structures; Google Scholar cited by 686 (Jul. 18, 2026); ESI Highly Cited Paper