Article
Geochemistry & Geophysics
Wenhua Zhang, Licheng Jiao, Fang Liu, Jia Liu, Zhen Cui
Summary: This research proposes a Laplacian high-frequency convolutional block (LHCB) based on convolutional neural networks (CNNs) to extract useful high-frequency features. By embedding LHCB into existing CNN structures, LHNet is obtained, and a high-frequency pathway (HFP) is introduced to propagate the blurred residual high-frequency features. Furthermore, a new objective function is proposed to enhance the intraclass similarity of high-frequency features. Experimental results demonstrate the superior performance of the proposed method in remote sensing image scene classification.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Engineering, Electrical & Electronic
Duong Hai Nguyen, Se-Ho Lee, Chul Lee
Summary: In this letter, a multiscale coarse-to-fine guided screenshot demoireing algorithm is proposed. Multiscale features of the input image are extracted, followed by the development of the multiscale guided restoration block (MGRB), which removes moire patterns by utilizing the correlation between moire frequencies. Two blocks for feature modulation and moire pattern removal are designed, and an adaptive reconstruction loss is developed to improve performance by focusing on difficult regions to restore. Experimental results on multiple datasets demonstrate comparable or even better demoireing performance than state-of-the-art algorithms.
IEEE SIGNAL PROCESSING LETTERS
(2023)
Article
Environmental Sciences
Zhichao Yuan, Ziming Liu, Chunbo Zhu, Jing Qi, Danpei Zhao
Summary: The proposed Multi-Feature Pyramid Network (MFPNet) with Receptive Field Block (RFB) effectively addresses the challenges of object detection in optical remote sensing images. Experimental results on the Levir and DIOR datasets demonstrate that the method outperforms state-of-the-art networks, showcasing better performance in target detection.
Article
Neurosciences
Zhengyuan Xu, Wentao Xiang, Songsheng Zhu, Rui Zeng, Cesar Marquez-Chin, Zhen Chen, Xianqing Chen, Bin Liu, Jianqing Li
Summary: The proposed LatLRR-FCNs framework successfully achieves superior performance in four medical image fusion tasks by utilizing LatLRR and FCNs modules.
FRONTIERS IN NEUROSCIENCE
(2021)
Article
Environmental Sciences
Xianpeng Guo, Biao Hou, Zitong Wu, Bo Ren, Shuang Wang, Licheng Jiao
Summary: This study proposes a novel framework, Prob-POS, to improve the interpretability of CNNs on remote sensing images. The framework utilizes a probe network and a weighted probability of occlusion (wPO) selection strategy to generate elaborate explanation maps and automatically select the optimal explanation layer. Experimental results demonstrate that Prob-POS enhances the faithfulness and explainability of CNNs on remote sensing images.
Article
Computer Science, Information Systems
Ines Brusch
Summary: This paper explores how image data can be automatically analyzed using a combination of image analysis methods and fuzzy cluster algorithms to predict user preferences. Depending on the diversity of the images, either SVM or CNN provide the best basis for preference prediction.
INFORMATION & MANAGEMENT
(2022)
Article
Engineering, Electrical & Electronic
Yuping Huang, Weisheng Li, Jiao Du
Summary: Medical image fusion technology enhances the accuracy of clinical diagnosis and treatment by retaining information from original medical images. The proposed multi-modal medical image fusion framework utilizes various steps to achieve image fusion and demonstrates superior performance in experimental results.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2021)
Article
Geochemistry & Geophysics
Wenhua Zhang, Licheng Jiao, Yuxuan Li, Zhongjian Huang, Haoran Wang
Summary: The paper introduces a novel Laplacian Feature Pyramid Network (LFPN) that combines low-frequency and high-frequency features to enhance object detection performance in VHR-ORS images. High-frequency features, crucial for distinguishing ground objects, have not been adequately addressed in previous studies.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Engineering, Electrical & Electronic
Yi-Si Luo, Xi-Le Zhao, Tai-Xiang Jiang, Yu-Bang Zheng, Yi Chang
Summary: In this article, the spatial-spectral constrained deep image prior (S2DIP) method is proposed for HSI mixed noise removal, integrating deep image prior (DIP), spatial-spectral total variation regularization term, and l(1)-norm sparse term to capture deep prior and noise structures. The experimental results demonstrate that S2DIP outperforms other model-based and deep learning-based state-of-the-art HSI denoising methods in terms of denoising ability.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Environmental Sciences
Hongwei Guo, Hongyang Bai, Yuman Yuan, Weiwei Qin
Summary: A novel method for ship detection in high spatial resolution remote sensing imagery is proposed, which utilizes deformable convolution networks to effectively extract features at different scales and orientations. Experimental results on public remote sensing datasets demonstrate the effectiveness of the proposed method in accurately detecting ships in remote sensing applications.
Article
Computer Science, Artificial Intelligence
Juan Wang, Yiping Duan, Xiaoming Tao, Mai Xu, Jianhua Lu
Summary: This paper proposes a novel DNN-based image compression framework, decomposing images into sub-images using multi-scale networks and independently optimizing each pyramid scale to achieve better compression representation, achieving a good trade-off between rate, distortion and perception.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Computer Science, Information Systems
Ines Brusch
Summary: This paper demonstrates how image data can be automatically analyzed using a combination of image analysis methods and fuzzy cluster algorithms to predict user preferences, which can help companies make targeted offers.
INFORMATION & MANAGEMENT
(2022)
Article
Computer Science, Software Engineering
Fengyuan Zuo, Yongdong Huang, Qiufu Li, Weijian Su
Summary: This paper proposes a deep multi-scale pyramid network for infrared and visible image fusion. The proposed method efficiently extracts and fuses the multi-scale detail information of source images, preserving more texture details compared to previous state-of-the-art methods.
INTERNATIONAL JOURNAL OF WAVELETS MULTIRESOLUTION AND INFORMATION PROCESSING
(2022)
Article
Computer Science, Information Systems
Rui Tang, Lihui Chen, Rongzhu Zhang, Awais Ahmad, Marcelo Keese Albertini, Xiaomin Yang
Summary: This paper introduces a network named LDSRN that combines the Laplacian pyramid structure and dense connection for reconstructing high-resolution medical images. Experimental results show that LDSRN achieves better results compared to other methods in terms of objective indices and subjective evaluations.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Geochemistry & Geophysics
Haiwen Du, Yu An, Qing Ye, Jiulin Guo, Lu Liu, Dongjie Zhu, Conrad Childs, John Walsh, Ruihai Dong
Summary: Seismic interpretation is a fundamental method for obtaining information about subsurface reservoirs. However, deep learning algorithms often underperform on seismic data due to inconsistent noise patterns. To address this issue, a noise pattern transfer framework is proposed to improve the generality of seismic interpretation algorithms. Experimental results demonstrate the effectiveness of this approach.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Chemistry, Analytical
Jing Li, Weiguo Gong, Weihong Li
Article
Environmental Sciences
Yan Zhang, Weiguo Gong, Jingxi Sun, Weihong Li
Article
Computer Science, Artificial Intelligence
Wenqian Liu, Weihong Li, Weiguo Gong
IET COMPUTER VISION
(2020)
Article
Computer Science, Artificial Intelligence
Jinkai Cui, Weihong Li, Weiguo Gong
Article
Environmental Sciences
Wei Guo, Weihong Li, Weiguo Gong, Jinkai Cui
Article
Environmental Sciences
Yan Zhang, Weihong Li, Weiguo Gong, Zixu Wang, Jingxi Sun
Article
Engineering, Electrical & Electronic
Bingxin Zhao, Weihong Li, Weiguo Gong
Summary: Most existing motion deblurring methods rely on paired images for training, which limits their effectiveness in real-world scenarios. To address this, we propose an unsupervised real-aware motion deblurring method using multi-attention CycleGAN with contrastive guidance. The network architecture consists of two streams, each handling unpaired sharp and blurred images. We introduce multi-attention GANs to capture real-aware information, as well as gradient contrastive and adversarial contrastive loss functions to enhance network representation. Experimental results on benchmark datasets demonstrate the superiority of our method over current state-of-the-art approaches for unsupervised motion deblurring.
DIGITAL SIGNAL PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
Bingxin Zhao, Weihong Li
Summary: A domain translation network with contrastive constraint is proposed for unpaired motion image deblurring, which overcomes the limitation of requiring a large amount of paired training data. The network includes two streams, a sharp domain translation and a blurred domain translation, to handle unpaired sharp and blurred images from the real world. The network utilizes a contrastive constraint loss in the deep intermediate level to produce deblurring results close to the real sharp image, and distinct loss functions to preserve edge and texture detail. Extensive experiments demonstrate that the proposed network outperforms state-of-the-art methods for unpaired motion image deblurring.
IET IMAGE PROCESSING
(2023)
Article
Environmental Sciences
Wei Guo, Weihong Li, Zhenghao Li, Weiguo Gong, Jinkai Cui, Xinran Wang
Article
Computer Science, Information Systems
Bingxin Zhao, Weihong Li, Weiguo Gong
Proceedings Paper
Remote Sensing
Jingxi Sun, Weihong Li, Yan Zhang, Weiguo Gong
IMAGE AND SIGNAL PROCESSING FOR REMOTE SENSING XXV
(2019)
Proceedings Paper
Engineering, Electrical & Electronic
Ying Liao, Weihong Li, Jinkai Cui, Weiguo Gong
2018 INTERNATIONAL CONFERENCE ON DIGITAL IMAGE COMPUTING: TECHNIQUES AND APPLICATIONS (DICTA)
(2018)
Proceedings Paper
Computer Science, Artificial Intelligence
Weiguo Gong, Xi Chen, Jinming Li, Yongliang Tang, Weihong Li
ADVANCES IN NEURAL NETWORKS, PT II
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Weihong Li, Yangqing Chen, Rui Chen, Weiguo Gong, Bingxin Zhao
ADVANCES IN NEURAL NETWORKS, PT II
(2017)