Article
Computer Science, Artificial Intelligence
Sen Jia, Zhichao Min, Xiyou Fu
Summary: In this paper, a Multiscale Spatial-spectral Transformer Network (MSST-Net) is proposed to extract spectral and spatial features from HSI and MSI using the self-attention mechanism of the Transformer. A self-supervised pre-training strategy is also introduced to improve the network's performance. Experimental results demonstrate that the proposed network achieves better performance compared to other state-of-the-art fusion methods.
INFORMATION FUSION
(2023)
Article
Environmental Sciences
Hongyu Zhao, Kaiyuan Feng, Yue Wu, Maoguo Gong
Summary: This paper proposes a novel feature extraction network that combines Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN) for hyperspectral change detection tasks. The experimental results demonstrate that the proposed method yields reliable detection results and has fewer noise regions.
Article
Geochemistry & Geophysics
Xiyou Fu, Sen Jia, Meng Xu, Jun Zhou, Qingquan Li
Summary: In this letter, a novel sparsity constrained fusion method based on matrix factorization is proposed for fusing hyperspectral and multispectral images. By imposing l(1) norm constraint and inserting a prior, this method can effectively handle localized changes between multiplatform images.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Engineering, Electrical & Electronic
Hao Shi, Guo Cao, Youqiang Zhang, Zixian Ge, Yanbo Liu, Di Yang
Summary: In the field of hyperspectral image (HSI) classification, recent advances have been made in deep learning-based research, particularly in convolutional neural network (CNN)-based classification and transformer-based classification. Both methodologies focus on exchanging information locally or at a long distance for HSI pixels in the spatial or spectral-spatial domain. However, there is still room for improvement in terms of accuracy and efficiency. To address this, fast Fourier transform (FFT) is introduced to enhance information mixing in HSI classification, resulting in a novel deep neural network architecture called fast Fourier filter network (F3 Net). Experimental results demonstrate the competitiveness of F3 Net, especially in situations with limited training samples.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Computer Science, Artificial Intelligence
Zhengjue Wang, Bo Chen, Hao Zhang, Hongwei Liu
Summary: This study proposes a nonlinear variational probabilistic generative model (NVPGM) based on nonlinear unmixing for unsupervised fusion of high-resolution hyperspectral images, using neural networks to implement nonlinear functions. By inferring latent representations with recognition models and using stochastic gradient variational inference, both latent representations and parameters can be simultaneously inferred to retrieve the target HR-HSI via feedforward mapping.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Geochemistry & Geophysics
Xueting Zhang, Wei Huang, Qi Wang, Xuelong Li
Summary: The article proposes an interpretable spatial-spectral reconstruction network (SSR-NET) based on CNN for efficient fusion of HSI and MSI. The SSR-NET consists of three components for cross-mode message inserting, spatial reconstruction, and spectral reconstruction, achieving superior or competitive results in comparison with seven state-of-the-art methods on six HSI data sets.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2021)
Article
Geochemistry & Geophysics
Yujuan Guo, Xiyou Fu, Meng Xu, Sen Jia
Summary: This research proposes an efficient method for fusing unregistered low-resolution hyperspectral images and high-resolution multispectral images. By using pixel shifting, the method achieves high-resolution, high signal-to-noise ratio, and feature identifiability in the fused images. The research also introduces a simple and stackable fusion block and a stereo cross-attention network for accurate fusion.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Environmental Sciences
Ping Zhang, Haiyang Yu, Pengao Li, Ruili Wang
Summary: This paper proposes a novel method called TransHSI for hyperspectral image classification, which leverages 3D CNNs and Transformer blocks to extract both spectral and spatial features. The fusion module combines shallow and deep features and applies a semantic tokenizer to enhance feature discriminability. Experimental results show that TransHSI achieves competitive performance on multiple datasets.
Article
Engineering, Electrical & Electronic
Vinod Kumar, Ravi Shankar Singh, Yaman Dua
Summary: The use of hyperspectral images for land cover mapping is an important research topic, with methods like CNN providing good classification results. The proposed MDCNN model combines mathematical morphology and convolutional neural networks to extract more robust spectral-spatial features, showing better classification results than traditional deep learning models.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2022)
Article
Geochemistry & Geophysics
Xiyou Fu, Sen Jia, Meng Xu, Jun Zhou, Qingquan Li
Summary: This article proposes a novel group sparsity constrained fusion method based on matrix factorization for fusing hyperspectral and multispectral images. By considering localized interimage changes and integrating an advanced denoiser, this method achieves better fusion results.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Environmental Sciences
Lijuan Su, Yuxiao Sui, Yan Yuan
Summary: In this paper, a GAN-based unsupervised HSI-MSI fusion network is proposed to reconstruct high resolution hyperspectral images. Experimental results show that the proposed method outperforms the state-of-art methods.
Article
Computer Science, Artificial Intelligence
Haifeng Sima, Feng Gao, Yudong Zhang, Junding Sun, Ping Guo
Summary: In this paper, a collaborative optimization parallel convolution network consisting of 3D-2D CNN is proposed for accurate classification of hyperspectral images. The experimental results show that this method outperforms the state-of-the-art methods and has better generalization capability.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2023)
Article
Environmental Sciences
Zhihui Wang, Baisong Cao, Jun Liu
Summary: This paper proposes a spatial shuffle strategy that selects a smaller neighborhood window and randomly shuffles the pixels within the window, aiming to simulate the potential patterns of the pixel distribution in the real world as much as possible. Experimental results show that smaller neighborhood windows can achieve the same or even better classification performance compared to larger neighborhood windows.
Article
Geochemistry & Geophysics
Wei-Ye Wang, Heng-Chao Li, Yang-Jun Deng, Li-Yang Shao, Xiao-Qiang Lu, Qian Du
Summary: This study introduces a novel generative adversarial network (GAN) for hyperspectral image (HSI) classification, which uses artificial sample generation for data augmentation to improve classification performance. Experimental results show that the proposed method outperforms several state-of-the-art deep classification methods on two real HSI datasets.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2021)
Article
Geochemistry & Geophysics
Tatiana Gelvez-Barrera, Henry Arguello, Alessandro Foi
Summary: The study presents a method for fusing high-spatial-and-low-spectral resolution multispectral image (MSI) with a low-spatial-and-high-spectral resolution hyperspectral image (HSI) to generate a high-resolution image (HRI). The proposed method uses various low-rank regularizations jointly and incorporates a nonlocal patch-based denoiser in the alternating direction method of multipliers (ADMM) to refine the spatial and spectral correlations of the HRI from the individual HSI and MSI data. The method outperforms state-of-the-art methods in recovering low-contrast areas and introduces a rank-one similarity prior, which is found to be an inherent characteristic of the HRI.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Editorial Material
Computer Science, Artificial Intelligence
Jun Zhou, Fengchao Xiong, Lei Tong, Naoto Yokoya, Pedram Ghamisi
IET COMPUTER VISION
(2023)
Article
Computer Science, Information Systems
Yaoming Cai, Zijia Zhang, Pedram Ghamisi, Behnood Rasti, Xiaobo Liu, Zhihua Cai
Summary: The article introduces a new method for clustering multimodal remote sensing data, which achieves state-of-the-art performance on large-scale multimodal datasets by employing Transformer, online clustering mechanism, and self-supervised training strategy, and has good scalability.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Xian Li, Mingli Ding, Yanfeng Gu, Aleksandra Pizurica
Summary: This paper introduces a unified deep learning framework for joint denoising and classification of high-dimensional images, particularly in the framework of hyperspectral imaging. The proposed joint learning framework substantially improves the classification performance and enhances the denoising results, especially in terms of the semantic content, benefiting from the classification.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Shengyang Li, Xian Sun, Yanfeng Gu, Yixuan Lv, Manqi Zhao, Zhuang Zhou, Weilong Guo, Yuhan Sun, Han Wang, Jian Yang
Summary: Intelligent processing of satellite video focuses on extracting specific information of ground objects and scenes from earth observation videos through intelligent image/video processing technology. This article presents a systematic review and quantitative analysis of the results published over the last decade, intending to further promote the development of various intelligent processing tasks for satellite video. It analyzes the current difficulties, challenges, and methodological systems for each task, and also provides in-depth analysis and summary of publicly available datasets, algorithm performance, and application scenarios.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Dakuan Du, Yanfeng Gu, Tianzhu Liu, Xian Li
Summary: In this article, a novel convolution and transformer joint network (CTJN) is proposed to address the challenge of high-accuracy spectral reconstruction (SR) in complex scenes. The CTJN utilizes shallow feature extraction modules (SFEMs) and deep feature extraction modules (DFEMs) to explore local spatial features and global spectral features. Additionally, a high-frequency transformer block (HF-TB) is designed to preserve detailed features and a spatial-spectral recalibration block (SSRB) is incorporated to enforce explicit constraints. Experimental results on multiple datasets demonstrate the superior performance of CTJN compared to state-of-the-art methods in both large- and small-scale scenes.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Lingbo Huang, Yushi Chen, Xin He
Summary: In this study, a spectral-spatial masked Transformer (SS-MTr) is explored for hyperspectral image (HSI) classification. A two-stage training strategy is utilized, where the Transformer is pretrained via reconstruction and then fine-tuned. Additionally, three SS-MTr-based methods are proposed to incorporate discriminative feature learning. Experimental results demonstrate that these methods achieve competitive performance compared to state-of-the-art HSI classification methods.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Xin He, Yushi Chen, Lingbo Huang
Summary: In recent years, deep learning models have been widely used for hyperspectral image (HSI) classification. However, most existing methods only focus on high classification accuracy and ignore uncertainty. To address this issue, Bayesian deep learning (BDL) is investigated to analyze the model uncertainty for HSI classification. Experimental results on public HSI datasets demonstrate the superiority of the proposed BDL-based methods in both accuracy and uncertainty estimation.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Chen Wang, Yanfeng Gu, Xian Li
Summary: In this article, a robust 3-D reconstruction method for multispectral images is proposed, which improves the accuracy and performance of reconstruction by using reflectance correction, band alignment, and multispectral feature extraction.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Meiqi Wang, Junfang Yang, Shanwei Liu, Yanfeng Gu, Mingming Xu, Yi Ma, Jie Zhang, Jianhua Wan
Summary: In this study, a 1DCNN_GRU model was developed to quantitatively invert the oil film thickness (OFT) by analyzing spectral characteristics. Experimental results showed that the proposed model effectively addressed the issue of poor spectral separability and outperformed other models in terms of inversion accuracy. Additionally, airborne hyperspectral data performed well in OFT inversion, especially in certain ranges, and the use of brightness temperature data improved the inversion accuracy.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Zhe Dong, Tianzhu Liu, Yanfeng Gu
Summary: This article proposes a spatial and semantic consistency self-supervised contrastive learning (SSCCL) framework for remote sensing semantic segmentation tasks. By integrating a consistency branch and an instance branch, the framework can learn robust and informative feature representations in limited annotated scenarios, achieving superior performance compared to state-of-the-art CL methods and ImageNet pretraining.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Likun Chen, Yanfeng Gu, Xian Li, Xiangrong Zhang, Baisen Liu
Summary: An article proposes a normalized spatial-spectral supervoxel segmentation method for multispectral point cloud (MPC) data, which can segment MPC without the need for any manual annotation and achieves better performance compared to other methods, as demonstrated in experiments.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Qingwang Wang, Cheng Yin, Haochen Song, Tao Shen, Yanfeng Gu
Summary: In this study, a novel uncertainty-guided trustworthy fusion network (UTFNet) is proposed for RGB-T semantic segmentation. The uncertainty of each modality is estimated and used to guide the information fusion, resulting in improved accuracy, robustness, and trustworthiness of the segmentation model.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Geochemistry & Geophysics
Bin Guo, Tianzhu Liu, Yanfeng Gu
Summary: In this article, a novel structure-preserving discriminative distribution adaptive MS-HS image collaborative classification method is proposed to improve the classification accuracy of large-scene MS images. The method maximizes the distance between different classes by combining statistical properties and geometric constraints, and adaptively maps multiscale spectral-spatial features of MS-HS images to subspaces for classification. Experimental results on three sets of MS-HS datasets demonstrate the effectiveness of the proposed method in reducing the differences between MS-HS data and achieving better classification results.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Computer Science, Artificial Intelligence
Yonghao Xu, Weikang Yu, Pedram Ghamisi, Michael Kopp, Sepp Hochreiter
Summary: The text introduces a novel text-to-image generation network Txt2Img-MHN, which achieves realistic remote sensing image generation from text descriptions through hierarchical prototype learning. The method performs well in zero-shot classification, providing a good metric for evaluating the realism and semantic consistency of generated images.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2023)
Article
Geochemistry & Geophysics
Yanfeng Gu, Yanyuan Huang, Tianzhu Liu
Summary: Traditional spectral unmixing of satellite hyperspectral images faces challenges of severe spectral mixing and variability caused by external factors. In this study, a novel UAV-satellite spectral unmixing model with intrinsic image decomposition is proposed to address these problems. Experimental results show that the proposed method effectively improves the robustness and accuracy of the unmixing results.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
