Multi-Band Image Fusion Based on Spectral Unmixing
University of Cambridge(剑桥大学);Instituto de Telecomunicações(电信研究所);Instituto Superior Técnico(高等技术学院);Universidade de Lisboa(里斯本大学);IRIT/INP-ENSEEIHT(图卢兹国立高等电子技术、电子、计算机、水力学与电信学院);University of Toulouse(图卢兹大学);Nanyang Technological University(南洋理工大学)
Qi Wei, Jose Bioucas-Dias, Nicolas Dobigeon, Jean-Yves Tourneret, Marcus Chen, Simon Godsill
This paper presents a multi-band image fusion algorithm based on unsupervised spectral unmixing for combining a high-spatial low-spectral resolution image and a low-spatial high-spectral resolution image. The widely used linear observation model (with additive Gaussian noise) is combined with the linear spectral mixture model to form the likelihoods of the observations. The non-negativity and sum-to-one constraints resulting from the intrinsic physical properties of the abundances are introduced as prior information to regularize this ill-posed problem. The joint fusion and unmixing problem is then formulated as maximizing the joint posterior distribution with respect to the endmember signatures and abundance maps, This optimization problem is attacked with an alternating optimization strategy. The two resulting sub-problems are convex and are solved efficiently using the alternating direction method of multipliers. Experiments are conducted for both synthetic and semi-real data. Simulation results show that the proposed unmixing based fusion scheme improves both the abundance and endmember estimation comparing with the state-of-the-art joint fusion and unmixing algorithms.
Band selection in RKHS for fast nonlinear unmixing of hyperspectral images
Federal University of Santa Catarina(圣卡塔琳娜联邦大学);Université de Nice Sophia-Antipolis(尼斯索菲亚-安蒂波利斯大学);CNRS(法国国家科学研究中心);University of Toulouse(图卢兹大学);IRIT-ENSEEIHT
Technical Report: Band selection for nonlinear unmixing of hyperspectral images as a maximal clique problem
Federal University of Santa Catarina(圣卡塔琳娜联邦大学);University of Nice Sophia-Antipolis(尼斯索菲亚-安蒂波利斯大学);Lagrange Laboratory (CNRS, OCA)(拉格朗日实验室(法国国家科学研究中心、蔚蓝海岸天文台));INRIA Sophia-Antipolis(法国国家信息与自动化研究所索菲亚-安蒂波利斯中心)
WS-Net: Weak-Signal Representation Learning and Gated Abundance Reconstruction for Hyperspectral Unmixing via State-Space and Weak Signal Attention Fusion
School of Information and Communication Technology, Griffith University(信息与通信技术学院,格里菲斯大学);School of Computing, Engineering and Mathematical Sciences, La Trobe University(计算、工程与数学科学学院,拉特罗布大学);CSIRO Agriculture and Food(澳大利亚农业与食品CSIRO)
R-FUSE: Robust Fast Fusion of Multi-Band Images Based on Solving a Sylvester Equation
University of Cambridge(剑桥大学);University of Toulouse(图卢兹大学);Instituto de Telecomunicações(电信研究所);Instituto Superior Técnico(高等技术学院);Universidade de Lisboa(里斯本大学)
Decoupled-and-Coupled Networks: Self-Supervised Hyperspectral Image Super-Resolution with Subpixel Fusion
Aerospace Information Research Institute, Chinese Academy of Sciences(中国科学院空天信息创新研究院);The University of Tokyo(东京大学);RIKEN(日本理化学研究所);Univ. Grenoble Alpes(格勒诺布尔阿尔卑斯大学);CNRS(法国国家科学研究中心);Grenoble INP(格勒诺布尔理工学院)
Image Fusion and Re-Modified SPIHT for Fused Image
本文提出了基于离散小波的融合技术,结合感知重要的图像特征。SPIHT(Set Partitioning in Hierarchical Trees)算法是一种有效的融合图像无损和有损编码方法。本文对SPIHT算法作了一些改进。该方法基于中高频子带间小波系数相关性不显著的思想。在RE-MSPIHT算法中,小波系数在SPIHT编码之前根据子带重要性进行缩放,目标是最小化MSE。
Correntropy Maximization via ADMM - Application to Robust Hyperspectral Unmixing
Institut Charles Delaunay (CNRS), Université de Technologie de Troyes(特鲁瓦技术大学,夏尔·德洛内研究所(法国国家科学研究中心));School of Engineering and Physical Sciences, Heriot-Watt University(赫瑞瓦特大学,工程与物理科学学院);LITIS lab, Université de Rouen(鲁昂大学,LITIS实验室)
Holistic Dynamic Frequency Transformer for Image Fusion and Exposure Correction
School of Biomedical Engineering and Instrument Science, Zhejiang University(浙江大学生物医学工程与仪器科学学院);School of Software Technology, Dalian University of Technology(大连理工大学软件学院);School of Mechanical Engineering, Dalian University of Technology(大连理工大学机械工程学院)
Hyperspectral Image Fusion with Spectral-Band and Fusion-Scale Agnosticism
University of Electronic Science and Technology of China, School of Mathematical Sciences, Chengdu, China(电子科技大学数学学院);University of Electronic Science and Technology of China, Multi-Hazard Early Warning Key Laboratory of Sichuan Province, Chengdu, China(电子科技大学四川省多灾种早期预警重点实验室);University of Electronic Science and Technology of China, School of Computer Science and Engineering (School of Cyber Security), Chengdu, China(电子科技大学计算机科学与工程学院(网络安全学院))
Coupled Convolutional Neural Network with Adaptive Response Function Learning for Unsupervised Hyperspectral Super-Resolution
Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences(中国科学院空天信息创新研究院数字地球科学重点实验室);Flemish Institute for Technological Research (VITO)(弗拉芒技术研究院);IMEC Research Group, Ghent University(根特大学IMEC研究组);Remote Sensing Technology Institute (IMF), German Aerospace Center (DLR)(德国航空航天中心遥感技术研究所)
Hierarchy-Aware and Channel-Adaptive Semantic Communication for Bandwidth-Limited Data Fusion
State Key Laboratory of Advanced Rail Autonomous Operation(先进轨道交通自主运行全国重点实验室);Beijing Jiaotong University(北京交通大学);School of Electronic and Information Engineering, Beijing Jiaotong University(北京交通大学电子与信息工程学院);Linköping University(林雪平大学);Singapore University of Technology and Design(新加坡科技设计大学)
Degrade is Upgrade: Learning Degradation for Low-light Image Enhancement
School of Computer Science, Wuhan University(武汉大学计算机学院);University of Central Florida(中佛罗里达大学);Wuhan Institute of Technology(武汉工程大学);National Tsing Hua University(清华大学(台湾))
弱光图像增强的目的是在保持图像视觉自然度的同时提高图像的可见度。不同于现有的方法往往直接完成重光照任务,忽略了保真度和自然度恢复,我们调查的内在退化和重光照的低光图像,同时细化的细节和颜色分两步。受彩色图像公式(漫射照明颜色加上环境照明颜色)的启发,我们首先估计来自低光输入的退化以模拟环境照明颜色的失真,然后细化内容以恢复漫射照明颜色的损失。为此,我们提出了一种新的降级到细化生成网络(DRGN)。它的显著特点可以概括为:1)一种新的两步生成网络,用于退化学习和内容细化。该方法不仅优于一步法,而且能够合成足够多的成对样本,有利于模型训练; 2)采用多分辨率融合网络,以多尺度协同的方式表示目标信息(退化或内容),更有效地解决复杂的解混问题。在增强任务和联合检测任务上的大量实验都验证了该方法的有效性和效率,分别以\textit{0.70dB on average and 3.18\% in mAP}的优势超过了SOTA。代码可在\url{https://github.com/kuijiang0802/DRGN}上获得。
SWAN: Self-supervised Wavelet Neural Network for Hyperspectral Image Unmixing
Info Edge India Ltd;Department of Artificial Intelligence and Machine Learning(人工智能与机器学习系);Acharya Institute of Technology(阿查里亚理工学院);Indian Institute of Information Technology Vadodara (IIITV)(瓦多拉印度信息科技学院)
Unmixing Optical Signals from Undersampled Volumetric Measurements by Filtering the Pixel Latent Variables
CERVO Brain Research Centre, Université Laval(拉瓦尔大学CERVO脑研究中心);Institute Intelligence and Data, Université Laval(拉瓦尔大学智能与数据研究所);Centre for Optics, Photonics and Lasers, Université Laval(拉瓦尔大学光学、光子学与激光中心);MILA-Québec AI Institute(MILA-魁北克人工智能研究所)
Band Prompting Aided SAR and Multi-Spectral Data Fusion Framework for Local Climate Zone Classification
College of Computer Science and Technology, Harbin Engineering University(哈尔滨工程大学计算机科学与技术学院);School of Astronautics, Beihang University(北京航空航天大学宇航学院);School of Software, Dalian University of Technology(大连理工大学软件学院);State Key Laboratory of Space-Ground Integrated Information Technology(天地一体化信息技术国家重点实验室)
AI-Driven HSI: Multimodality, Fusion, Challenges, and the Deep Learning Revolution
School of Electrical Engineering and Computer Science, Gwangju Institute of Science and Technology (GIST)(光州科学技术院电气工程与计算机科学学院);Artificial Intelligence Graduate School, GIST(光州科学技术院人工智能研究生院)
Spectral Superresolution of Multispectral Imagery with Joint Sparse and Low-Rank Learning
Aerospace Information Research Institute, Chinese Academy of Sciences(中国科学院空天信息创新研究院);Univ. Grenoble Alpes(格勒诺布尔阿尔卑斯大学);CNRS(法国国家科学研究中心);Grenoble INP(格勒诺布尔理工学院);University of Chinese Academy of Sciences(中国科学院大学);University of Pavia(帕维亚大学);INRIA(法国国家信息与自动化研究所)
Euclidean Distance to Convex Polyhedra and Application to Class Representation in Spectral Images
Center for Mathematical Morphology (CMM), Mines Paris, PSL University(巴黎矿业学院 数学形态学中心(CMM),PSL大学);Laboratoire de Réactivité et Chimie des Solides (LRCS), CNRS UMR 7314, Université de Picardie Jules Verne(皮卡第儒勒·凡尔纳大学 固体反应性与化学实验室(LRCS),法国国家科研中心UMR 7314)
LiDAR-Guided Cross-Attention Fusion for Hyperspectral Band Selection and Image Classification
School of Information and Communication Technology, Griffith University(格里菲斯大学信息与通信技术学院);Queensland Department of Agriculture and Fisheries(昆士兰州农业与渔业厅)
提出了一种多波段遥感图像的贝叶斯融合方法。所观察到的图像与要通过物理退化恢复的高光谱和高空间分辨率图像相关,由传感器特性定义的空间和光谱模糊和/或子采样。在贝叶斯估计框架内制定的融合问题。一个适当的先验分布,利用几何考虑。为了从感兴趣场景的后验分布计算其贝叶斯估计,设计了马尔可夫链蒙特卡罗算法来根据目标分布生成渐近分布的样本。为了有效地从这种高维分布中采样,在Gibbs采样策略中引入了Hamilton Monte Carlo步骤。相对于几个国家的最先进的融合技术,所提出的融合方法的效率进行评估。特别地,低空间分辨率的高光谱和多光谱图像被融合以产生高空间分辨率的高光谱图像。
LGBP-OrgaNet: Learnable Gaussian Band Pass Fusion of CNN and Transformer Features for Robust Organoid Segmentation and Tracking
School of Optoelectronic Science and Engineering(光电科学与工程学院);Translational Chinese Medicine Key Laboratory of Sichuan, Sichuan-Chongqing Joint Key Laboratory of innovation of New Drugs of Traditional, Chinese Medicine Sichuan Institute for Translational Chinese Medicine(四川 translational 中医关键实验室,四川-重庆联合新药创新中药关键实验室,四川 translational 中医研究所);Sichuan Cancer Hospital and Institute(四川癌症医院与研究所)
Complementarity-Supervised Spectral-Band Routing for Multimodal Emotion Recognition
Great Bay University(大湾大学);The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳));City University of Hong Kong (Dongguan)(香港城市大学(东莞));Tencent Youtu Lab(腾讯优图实验室)