IEEE Journal of Selected Topics in Signal Processing最新文献_第6页

Unlocking Off-the-Grid Sparse Recovery With Unlimited Sensing: Simultaneous Super-Resolution in Time and Amplitude 无限传感解锁离网稀疏恢复：同时超分辨率的时间和幅度

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-10-27 DOI: 10.1109/JSTSP.2025.3626267

Ruiming Guo;Ayush Bhandari

{"title":"Unlocking Off-the-Grid Sparse Recovery With Unlimited Sensing: Simultaneous Super-Resolution in Time and Amplitude","authors":"Ruiming Guo;Ayush Bhandari","doi":"10.1109/JSTSP.2025.3626267","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3626267","url":null,"abstract":"The recovery of Dirac impulses, or spikes, from filtered measurements is a classical problem in signal processing. As the spikes lie in the continuous domain while measurements are discrete, this task is known as super-resolution or off-the-grid sparse recovery. Despite significant theoretical and algorithmic advances over the past decade, these developments often overlook critical challenges at the analog–digital interface. In particular, when spikes exhibit strong-weak amplitude disparity, conventional digital acquisition may result in clipping of strong components or loss of weak ones beneath the quantization noise floor. This motivates a broader perspective: super-resolution must simultaneously resolve both amplitude and temporal structure. Under a fixed bit budget, such information loss is unavoidable. In contrast, the emerging theory and practice of the Unlimited Sensing Framework (USF) demonstrate that these fundamental limitations can be overcome. Building on this foundation, we demonstrate that modulo encoding within USF enables digital super-resolution by enhancing measurement precision, thereby unlocking temporal super-resolution beyond conventional limits. We develop new theoretical results that extend to non-bandlimited kernels commonly encountered in practice and introduce a robust algorithm for off-the-grid sparse recovery. To demonstrate practical impact, we instantiate our framework in the context of time-of-flight imaging. Both numerical simulations and hardware experiments validate the effectiveness of our approach under low-bit quantization, enabling super-resolution in amplitude and time.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1808-1821"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148504280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Incremental 3D Near-Field SAR Imaging With Block-Sparse Recovery 基于块稀疏恢复的增量三维近场SAR成像

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-10-27 DOI: 10.1109/JSTSP.2025.3626263

Zhichao Sheng;Yechao Peng;Wei Hong;Ye Shi;Dan Zeng

{"title":"Incremental 3D Near-Field SAR Imaging With Block-Sparse Recovery","authors":"Zhichao Sheng;Yechao Peng;Wei Hong;Ye Shi;Dan Zeng","doi":"10.1109/JSTSP.2025.3626263","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3626263","url":null,"abstract":"Due to the sparsity of scattered signals, the theory of compressed sensing (CS) has natural applicability in the field of millimetre-wave (MMW) near-field synthetic aperture radar (SAR) imaging. However, the sparse basis matrix in conventional sparse imaging model not only has a high demand for storage resources, but also destroys the block-sparse structure of the SAR image, which leads to the difficulties in applying CS based imaging algorithms to the practical system and hinders relevant theoretical research to some extent. To overcome these problems, a sparse imaging model based on the discrete Fourier transform basis is proposed, which significantly reduces the size of sensing matrix and improves its generation speed while preserving the block-sparse structure, allowing better-performing block-sparse recovery algorithms to be employed for imaging. In order to design the corresponding sparse imaging algorithm and provide a reference for others, the generalized approximate message passing based pattern-coupled sparse Bayesian learning (GAMP-PCSBL), which can achieve satisfactory sparse recovery results through neighborhood association and approximate computation of the posterior distribution, is introduced with some modifications. Additionally, a generalized incremental 3D imaging framework based on the model is proposed, enabling efficient 3D imaging. Experiments based on public data sets demonstrate the effectiveness of proposed imaging model, algorithms and framework. Only approximately 6<inline-formula><tex-math>$mathbf{%}$</tex-math></inline-formula> and 14<inline-formula><tex-math>$mathbf{%}$</tex-math></inline-formula> of the entire data are required for single-target and multi-target imaging respectively.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1796-1807"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148504301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Photometric Regularization for 3D Gaussian Splatting in Multi-View Surface Projection 多视点表面投影中三维高斯飞溅的光度正则化

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-10-03 DOI: 10.1109/JSTSP.2025.3617861

Yakun Ju;Yuying Zhao;Jun Xiao;Cong Zhang;Zheheng Jiang;Huiyu Zhou;Wei Zhou;Hui Yu;Junyu Dong

{"title":"Photometric Regularization for 3D Gaussian Splatting in Multi-View Surface Projection","authors":"Yakun Ju;Yuying Zhao;Jun Xiao;Cong Zhang;Zheheng Jiang;Huiyu Zhou;Wei Zhou;Hui Yu;Junyu Dong","doi":"10.1109/JSTSP.2025.3617861","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3617861","url":null,"abstract":"3D surface representation from multi-view photometric stereo (MVPS) remains a challenging task due to accumulated errors and geometric inconsistencies in conventional fusion-based methods. While 3D Gaussian Splatting (3DGS) enables real-time rendering, its explicit point-based representation struggles to preserve fine geometric details when directly applied to MVPS. To address this limitation, we propose PS-GS, a novel photometric-regularized Gaussian splatting framework that integrates surface normal priors to enhance multi-view surface projection. Specifically, PS-GS introduces three regularization strategies: (1) normal-guided initialization, which aligns Gaussian splats with photometric surface normals; (2) adaptive splat cloning, which dynamically adjusts point density in regions with high normal variation to improve local detail preservation; and (3) visibility-aware occlusion handling, which ensures geometric consistency across multi-view projections. Unlike NeRF-based MVPS methods, our approach does not rely on implicit volumetric representations and instead optimizes explicit surface projections, achieving efficient and artifact-free rendering. Experimental results on the DiLiGenT-MV benchmark demonstrate that PS-GS significantly improves 2D photometric fidelity, achieving a 2.18 dB PSNR gain over NeRF-based MVPS methods while maintaining around 150× faster rendering efficiency. Additionally, our normal-guided splatting enhances view-consistent normal alignment, reducing artifacts in challenging regions. Although PS-GS does not perform full 3D mesh reconstruction or relighting, its efficient rendering and surface-aware regularization make it a promising approach for real-time multi-view image synthesis in MVPS applications.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1682-1693"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148504686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Mask-Informed Spatial-Spectral Attention Unfolding Network for Snapshot Compressive Reconstruction 基于掩模的空间光谱注意力展开网络用于快照压缩重构

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-11-03 DOI: 10.1109/JSTSP.2025.3628534

Tong Zhang;Wenxue Cui;Yuhang Qin;Haiqi Zhu;Shaohui Liu;Feng Jiang

{"title":"Mask-Informed Spatial-Spectral Attention Unfolding Network for Snapshot Compressive Reconstruction","authors":"Tong Zhang;Wenxue Cui;Yuhang Qin;Haiqi Zhu;Shaohui Liu;Feng Jiang","doi":"10.1109/JSTSP.2025.3628534","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3628534","url":null,"abstract":"Snapshot compressive imaging (SCI) facilitates the computational recovery of high-dimensional spatio-temporal data through single-shot 2D coded measurements. However, existing SCI reconstruction paradigms usually exhibit inherent limitations in comprehensively exploiting the structural information embedded in optical modulation masks while effectively modeling complex spatial-spectral interdependencies, resulting in suboptimal reconstruction accuracy. To address the above challenges, a novel Mask-Informed Spatial-Spectral Attention Unfolding Network (MSS-Net) is proposed, which is able to systematically explore optical mask priors and model spatial-spectral correlations within a deep unfolding architecture. Specifically, for mask prior exploration, the mask is comprehensively utilized throughout the entire unrolled iterative architecture: a) In the gradient descent process, the mask is employed to jointly generate dynamic gradient and step-size maps, enabling mask-guided fine-grained adaptive gradient updating. b) In the proximal mapping process, the mask provides rich contextual guidance for feature extraction at different scale spaces, facilitating effective mask information fusion and efficient learning of deep representations. On the other hand, for spatial-spectral correlation modeling, an adaptive feature fusion mechanism with a hierarchical encoder-decoder architecture is proposed, which is able to jointly capture fine-grained spatial details and long-range spectral dependencies, thus ensuring high structural fidelity and robust spectral coherence in the reconstructed data. Extensive experiments demonstrate that the proposed MSS-Net outperforms the existing SCI reconstruction approaches on both simulated and real HSI datasets.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1850-1860"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148512611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

VTGF-IR: Visible Texture-Guided-Free Infrared Imagery Reconstruction VTGF-IR：可见纹理引导的无红外图像重建

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-07-31 DOI: 10.1109/JSTSP.2025.3594246

Pengcheng Jin;Qingwang Wang;Tao Shen

{"title":"VTGF-IR: Visible Texture-Guided-Free Infrared Imagery Reconstruction","authors":"Pengcheng Jin;Qingwang Wang;Tao Shen","doi":"10.1109/JSTSP.2025.3594246","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3594246","url":null,"abstract":"Thermal infrared images exhibit deficient texture details due to the inherent limitations of sensors. Existing infrared image enhancement methods remain constrained by insufficient analysis of thermal imaging principles and the intrinsic entropy bottleneck of texture information, resulting in suboptimal performance. To address this challenge, we propose the Visible Texture-Guided-Free Infrared Imagery Reconstruction (VTGF-IR) framework. Specifically, we develop a thermal radiation reconstruction module based on infrared imaging principles to suppress noise interference. To introduce additional texture information into infrared images while enhancing texture representation capability, we design a visible-infrared texture representation transfer module that utilizes visible light texture encoding patterns to guide infrared image texture reconstruction. Furthermore, we co-optimize texture representation through combined reconstruction loss and texture transfer loss. By integrating thermal imaging principles with visible-light guidance, our framework provides a hardware-independent solution for high-quality infrared image reconstruction. Experimental results demonstrate that the proposed method significantly enhances texture details in infrared images, achieving state-of-the-art performance across all evaluation metrics.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1632-1642"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148512615","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MoA-VR: A Mixture-of-Agents System Toward All-in-One Video Restoration MoA-VR：实现一体化视频还原的混合代理系统

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-11-04 DOI: 10.1109/JSTSP.2025.3627128

Lu Liu;Chunlei Cai;Shaocheng Shen;Jianfeng Liang;Weimin Ouyang;Tianxiao Ye;Jian Mao;Huiyu Duan;Jiangchao Yao;Xiaoyun Zhang;Qiang Hu;Guangtao Zhai

{"title":"MoA-VR: A Mixture-of-Agents System Toward All-in-One Video Restoration","authors":"Lu Liu;Chunlei Cai;Shaocheng Shen;Jianfeng Liang;Weimin Ouyang;Tianxiao Ye;Jian Mao;Huiyu Duan;Jiangchao Yao;Xiaoyun Zhang;Qiang Hu;Guangtao Zhai","doi":"10.1109/JSTSP.2025.3627128","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3627128","url":null,"abstract":"Real-world videos often suffer from complex degradations, such as noise, compression artifacts, and low-light distortions, due to diverse acquisition and transmission conditions. Existing restoration methods typically require professional manual selection of specialized models or rely on monolithic architectures that fail to generalize across varying degradations. Inspired by expert experience, we propose MoA-VR, the first <underline>Mixture-<underline>of-<underline>Agents <underline>Video <underline>Restoration system that mimics the reasoning and processing procedures of human professionals through three coordinated agents: Degradation Identification, Routing and Restoration, and Restoration Quality Assessment. Specifically, we construct a large-scale and high-resolution video degradation recognition benchmark and build a vision-language model (VLM) driven degradation identifier. We further introduce a self-adaptive router powered by large language models (LLMs), which autonomously learns effective restoration strategies by observing tool usage patterns. To assess intermediate and final processed video quality, we construct the <underline>Restored <underline>Video <underline>Quality (Res-VQ) dataset and design a dedicated VLM-based video quality assessment (VQA) model tailored for restoration tasks. Extensive experiments demonstrate that MoA-VR effectively handles diverse and compound degradations, consistently outperforming existing baselines in terms of both objective metrics and perceptual quality. These results highlight the potential of integrating multimodal intelligence and modular reasoning in general-purpose video restoration systems.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1822-1837"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148504281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Unified Framework for Deblurring and HDR Imaging in Dynamic Scenes 动态场景中去模糊和HDR成像的统一框架

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-06-19 DOI: 10.1109/JSTSP.2025.3581039

Xiaowen Ma;Kangbiao Shi;Daijin Chen;Yu Cao;Qingsen Yan;Yanning Zhang

{"title":"A Unified Framework for Deblurring and HDR Imaging in Dynamic Scenes","authors":"Xiaowen Ma;Kangbiao Shi;Daijin Chen;Yu Cao;Qingsen Yan;Yanning Zhang","doi":"10.1109/JSTSP.2025.3581039","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3581039","url":null,"abstract":"High Dynamic Range (HDR) imaging endeavors to enhance the visual appeal of an image by merging multi-exposure images. However, images captured with hand-held cameras often suffer from severe motion blur and ghosting artifacts in dynamic scenes. To address these challenges, we propose Image Deblurring and HDR Imaging (ID-HDRI), an end-to-end joint optimization framework that directly recovers sharp details and reconstructs HDR images from blurred, multi-exposure inputs. The proposed framework employs a dual-branch architecture that divides the encoder into two individual branches, each dedicated to a specific task. This innovative design facilitates the separate learning of degradation features. Furthermore, to fully utilize the potential of these features, we introduce multi-scale fusion and gate fusion modules, which play key roles in the entire network. Additionally, we build a new dataset comprising images with varying exposure levels and blurriness for HDR imaging in dynamic settings. Finally, ablation analyses demonstrate the efficacy of these fusion modules in enhancing HDR image reconstruction performance. Compared to traditional pipelines that sequentially perform CNN-based deblurring and HDR imaging, our method exhibits superior performance in both objective metrics and subjective visual quality.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1579-1588"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148512606","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Deep Physically Parameterized All-in-One Network for Lens-Free Microscopy Imaging 用于无透镜显微镜成像的深度物理参数化一体化网络

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-10-10 DOI: 10.1109/JSTSP.2025.3620707

Zhiyi Zhang;Yixiao Yang;Xudong Zhao;Ran Tao

{"title":"Deep Physically Parameterized All-in-One Network for Lens-Free Microscopy Imaging","authors":"Zhiyi Zhang;Yixiao Yang;Xudong Zhao;Ran Tao","doi":"10.1109/JSTSP.2025.3620707","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3620707","url":null,"abstract":"Lens-free microscopy circumvents the conventional trade-off between large field-of-view and high resolution by eliminating bulky optical lenses and relying on computational phase retrieval. However, existing reconstruction methods often face challenges in simultaneously ensuring fidelity, generalization, and efficiency, restricting their practical deployment. To address these limitations, we propose a deep physically parameterized all-in-one network for single-shot phase retrieval in lens-free microscopy. The framework leverages the fractional Fourier transform as a flexible forward model, where the fractional order and noise intensity are served as controllable parameters to characterize diverse diffraction imaging settings. Building on this formulation, we design an all-in-one reconstruction network that integrates an optimization operator, a controllable proximal network, and a controllable squeeze-and-excitation module, while explicitly embedding physical parameters to dynamically adapt the reconstruction process across varying imaging scenarios. Extensive simulations and experimental results demonstrate that the proposed method, using a single pre-trained model, generalizes effectively to diverse diffraction settings and consistently achieves state-of-the-art performance in both reconstruction accuracy and robustness.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1710-1723"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148504670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FusionRF: High-Fidelity Satellite Neural Radiance Fields From Multispectral and Panchromatic Acquisitions FusionRF：来自多光谱和全色采集的高保真卫星神经辐射场

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-11-03 DOI: 10.1109/JSTSP.2025.3628532

Michael Sprintson;Rama Chellappa;Cheng Peng

{"title":"FusionRF: High-Fidelity Satellite Neural Radiance Fields From Multispectral and Panchromatic Acquisitions","authors":"Michael Sprintson;Rama Chellappa;Cheng Peng","doi":"10.1109/JSTSP.2025.3628532","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3628532","url":null,"abstract":"We introduce FusionRF, a novel framework for digital surface reconstruction from satellite multispectral and panchromatic images. Current work has demonstrated the increased accuracy of neural photogrammetry for surface reconstruction from optical satellite images compared to algorithmic methods. Common satellites produce both a panchromatic and multispectral image, which contain high spatial and spectral information respectively. Current neural reconstruction methods require multispectral images to be upsampled with a pansharpening method using the spatial data in the panchromatic image. However, these methods may introduce biases and hallucinations due to domain gaps. FusionRF introduces joint image fusion during optimization through a novel cross-resolution kernel that learns to resolve spatial resolution loss present in multispectral images. As input, FusionRF accepts the original multispectral and panchromatic data, eliminating the need for image preprocessing. FusionRF also leverages multimodal appearance embeddings that encode the image characteristics of each modality and view within a uniform representation. By optimizing on both modalities, FusionRF learns to fuse image modalities while performing reconstruction tasks and eliminates the need for a pansharpening preprocessing step. We evaluate our method on multispectral and panchromatic satellite images from the WorldView-3 satellite in various locations, and show that FusionRF provides an average of 17% reduction in depth reconstruction error, and renders sharp training and novel views.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1838-1849"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148512610","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Diffusion-Driven Pansharpening via Conditional Multi-Stage Fusion 基于条件多阶段融合的扩散驱动泛锐化

IF 8.5 1区工程技术

IEEE Journal of Selected Topics in Signal Processing Pub Date : 2025-12-01 Epub Date: 2025-10-14 DOI: 10.1109/JSTSP.2025.3621350

Pengyu Chen;Libo Cheng;Liang-Jian Deng;Jun Liu;Jianing Sun;Tieyong Zeng

{"title":"Diffusion-Driven Pansharpening via Conditional Multi-Stage Fusion","authors":"Pengyu Chen;Libo Cheng;Liang-Jian Deng;Jun Liu;Jianing Sun;Tieyong Zeng","doi":"10.1109/JSTSP.2025.3621350","DOIUrl":"https://doi.org/10.1109/JSTSP.2025.3621350","url":null,"abstract":"Multispectral pansharpening aims to produce high-resolution multispectral (HRMS) images by fusing a low-resolution multispectral (LRMS) image with a corresponding panchromatic (PAN) image. While the rise of diffusion models has marked significant progress in this field, existing methods often struggle by directly feeding the PAN and LRMS images into the network, an operation that fails to provide sufficient guidance for generating high-quality HRMS images. To tackle this problem, we propose a novel multi-stage diffusion-driven model for pansharpening (MSDP) in this paper. Our method involves a three-stage conditional fusion mechanism where effective cross-modal prior fusion is achieved through a novel sparse attention module, with dynamic integration of PAN image shearlet coefficients to achieve multi-scale frequency enhancement. These conditions are injected into the model network via two modulation modules, ensuring the exceptional preservation of both spatial sharpness and spectral fidelity. Furthermore, a lightweight physical model is developed to accelerate the training process. Experimental evaluations on pansharpening datasets show that our proposed method achieves superior results compared to recent state-of-the-art (SOTA) techniques, highlighting its potential to advance the field of pansharpening.","PeriodicalId":13038,"journal":{"name":"IEEE Journal of Selected Topics in Signal Processing","volume":"19 8","pages":"1767-1781"},"PeriodicalIF":8.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148513687","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0