2024
2024
-
Record 349 of
Title:Thread the Needle: Cues-Driven Multiassociation for Remote Sensing Cross-Modal Retrieval
Author Full Names:Chen, Yaxiong; Huang, Jirui; Sun, Zhaoyang; Xiong, Shengwu; Lu, XiaoqiangSource Title:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSINGLanguage:EnglishDocument Type:ArticleKeywords Plus:IMAGE; TEXTAbstract:Rapid advances in Earth observation technologies have yielded numerous remotely sensed images and corresponding text data, enabling cross-modal image-text retrieval to extract valuable clues. However, current methods often focus on learning global semantic information from text and remote sensing (RS) images, while neglecting fine-grained semantic alignment and correlation. In addition, contrastive learning between modalities is often insufficient. To address these issues, we propose an innovative cues-driven multiassociation feature matching network (CDMAN) for cross-modal RS image retrieval. The proposed method primarily involves two key steps: 1) aligning positive samples and enhancing fusion for negative samples based on modal cues. To achieve precise alignment between RS images and text and facilitate the learning process for negative samples in contrastive learning, we have developed a novel fine-grained cues injection module that aligns and guides modalities using fine-grained cues; and 2) establishing multigranularity associative learning. To address the issue of insufficient association between RS images and text, we have implemented multigranularity collaborative associative learning, focusing on general and fine-grained modal associations. By fully leveraging modal cues, our method maintains both detailed associations and overall consistency in global associations. Experiments demonstrate that, compared to baseline methods, this approach achieves more accurate cross-modal retrieval (MCR) by combining fine-grained alignment and multigranularity associations.Addresses:[Chen, Yaxiong; Huang, Jirui; Sun, Zhaoyang] Wuhan Univ Technol, Sanya Sci & Educ Innovat Pk, Sanya 572000, Peoples R China; [Chen, Yaxiong; Huang, Jirui; Sun, Zhaoyang] Wuhan Univ Technol, Sch Comp Sci & Artificial Intelligence, Wuhan 430070, Peoples R China; [Chen, Yaxiong; Xiong, Shengwu] Interdisciplinary Artificial Intelligence Res Inst, Wuhan Coll, Wuhan 430212, Peoples R China; [Xiong, Shengwu] Shanghai Artificial Intelligence Lab, Shanghai 200232, Peoples R China; [Xiong, Shengwu] Qiongtai Normal Univ, Sch Informat Sci & Technol, Haikou 571127, Peoples R China; [Huang, Jirui; Sun, Zhaoyang] Wuhan Univ Technol, Chongqing Res Inst, Chongqing 401122, Peoples R China; [Lu, Xiaoqiang] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Key Lab Spectral Imaging Technol CAS, Xian 710119, Peoples R ChinaAffiliations:Wuhan University of Technology; Wuhan University of Technology; Wuhan College; Qiongtai Normal University; Wuhan University of Technology; Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CASPublication Year:2024Volume:62Article Number:4709813DOI Link:http://dx.doi.org/10.1109/TGRS.2024.3509639數(shù)據(jù)庫ID（收錄號）:WOS:001375996400029 -
Record 350 of
Title:One-Dimensional Gap Soliton Molecules and Clusters in Optical Lattice-Trapped Coherently Atomic Ensembles via Electromagnetically Induced Transparency
Author Full Names:Chen, Zhiming; Xie, Hongqiang; Zhou, Qi; Zeng, JianhuaSource Title:CRYSTALSLanguage:EnglishDocument Type:ArticleKeywords Plus:EQUATIONS; DYNAMICS; LIGHTAbstract:In past years, optical lattices have been demonstrated as an excellent platform for making, understanding, and controlling quantum matters at nonlinear and fundamental quantum levels. Shrinking experimental observations include matter-wave gap solitons created in ultracold quantum degenerate gases, such as Bose-Einstein condensates with repulsive interaction. In this paper, we theoretically and numerically study the formation of one-dimensional gap soliton molecules and clusters in ultracold coherent atom ensembles under electromagnetically induced transparency conditions and trapped by an optical lattice. In numerics, both linear stability analysis and direct perturbed simulations are combined to identify the stability and instability of the localized gap modes, stressing the wide stability region within the first finite gap. The results predicted here may be confirmed in ultracold atom experiments, providing detailed insight into the higher-order localized gap modes of ultracold bosonic atoms under the quantum coherent effect called electromagnetically induced transparency.Addresses:[Chen, Zhiming; Xie, Hongqiang; Zhou, Qi] East China Univ Technol, Sch Sci, Nanchang 330013, Peoples R China; [Chen, Zhiming; Zeng, Jianhua] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Ctr Attosecond Sci & Technol, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China; [Zeng, Jianhua] Univ Chinese Acad Sci, Sch Optoelect, Beijing 100049, Peoples R China; [Zeng, Jianhua] Shanxi Univ, Collaborat Innovat Ctr Extreme Opt, Taiyuan 030006, Peoples R ChinaAffiliations:East China University of Technology; State Key Laboratory of Transient Optics & Photonics; Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Chinese Academy of Sciences; University of Chinese Academy of Sciences, CAS; Shanxi UniversityPublication Year:2024Volume:14Issue:1Article Number:36DOI Link:http://dx.doi.org/10.3390/cryst14010036數(shù)據(jù)庫ID（收錄號）:WOS:001149031400001 -
Record 351 of
Title:Interface Contact Thermal Resistance of Die Attach in High-Power Laser Diode Packages
Author Full Names:Deng, Liting; Li, Te; Wang, Zhenfu; Zhang, Pu; Wu, Shunhua; Liu, Jiachen; Zhang, Junyue; Chen, Lang; Zhang, Jiachen; Huang, Weizhou; Zhang, RuiSource Title:ELECTRONICSLanguage:EnglishDocument Type:ArticleKeywords Plus:PERFORMANCEAbstract:The reliability of packaged laser diodes is heavily dependent on the quality of the die attach. Even a small void or delamination may result in a sudden increase in junction temperature, eventually leading to failure of the operation. The contact thermal resistance at the interface between the die attach and the heat sink plays a critical role in thermal management of high-power laser diode packages. This paper focuses on the investigation of interface contact thermal resistance of the die attach using thermal transient analysis. The structure function of the heat flow path in the T3ster thermal resistance testing experiment is utilized. By analyzing the structure function of the transient thermal characteristics, it was determined that interface thermal resistance between the chip and solder was 0.38 K/W, while the resistance between solder and heat sink was 0.36 K/W. The simulation and measurement results showed excellent agreement, indicating that it is possible to accurately predict the interface contact area of the die attach in the F-mount packaged single emitter laser diode. Additionally, the proportion of interface contact thermal resistance in the total package thermal resistance can be used to evaluate the quality of the die attach.Addresses:[Deng, Liting; Li, Te; Wang, Zhenfu; Zhang, Pu; Wu, Shunhua; Liu, Jiachen; Zhang, Junyue; Chen, Lang; Zhang, Jiachen; Huang, Weizhou; Zhang, Rui] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China; [Deng, Liting; Wu, Shunhua; Liu, Jiachen; Zhang, Junyue; Huang, Weizhou] Univ Chinese Acad Sci, Beijing 100049, Peoples R ChinaAffiliations:State Key Laboratory of Transient Optics & Photonics; Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Chinese Academy of Sciences; University of Chinese Academy of Sciences, CASPublication Year:2024Volume:13Issue:1Article Number:203DOI Link:http://dx.doi.org/10.3390/electronics13010203數(shù)據(jù)庫ID（收錄號）:WOS:001139159500001 -
Record 352 of
Title:GLGAT-CFSL: Global-Local Graph Attention Network-Based Cross-Domain Few-Shot Learning for Hyperspectral Image Classification
Author Full Names:Ding, Chen; Deng, Zhicong; Xu, Yaoyang; Zheng, Mengmeng; Zhang, Lei; Cao, Yu; Wei, Wei; Zhang, YanningSource Title:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSINGLanguage:EnglishDocument Type:ArticleKeywords Plus:CONVOLUTIONAL NETWORKS; ADAPTATIONAbstract:Few-shot learning (FSL) is an effective approach to address the issue of limited labeled data in hyperspectral image classification (HSIC). However, it overlooks the domain shift between the source domain (SD) and the target domain (TD) in cross-domain tasks. Most existing domain adaptation (DA) methods alleviate the domain shift problem to some extent, but DA methods based on traditional convolutional operators overlook the nonlocal spatial relationships in HSI, while methods based on graph neural networks (GNNs), although effective in leveraging nonlocal spatial information for domain alignment, overly emphasize global relationships, which is disadvantageous for pixel-level classification in HSI. To solve these issues, this article proposes a novel globalp-local graph attention network-based cross-domain FSL (GLGAT-CFSL), which comprehensively reduces domain shift through global-to-local domain alignment. It has the following advantages: 1) an innovative dynamic triplet graph attention network is devised to identify nonlocal spatial relationships in HSI for global graph alignment (GGA) while also addressing common overfitting and oversmoothing issues in GNNs; 2) an ingenious local similarity learning (LSL) strategy is designed after global domain alignment, utilizing intradomain connectivity structures and interdomain node similarities for local DA, promoting cross-domain information propagation and more comprehensive reduction of domain shift; and 3) we propose a novel triaxial dynamic convolutional neural network (TDCNN) as the feature extractor, promoting cross-dimensional interaction between spectral and spatial dimensions, establishing a more generalizable and rich feature representation between the SD and the TD. The experimental results on three HSI datasets demonstrate the superiority and effectiveness of the proposed GLGAT-CFSL.Addresses:[Ding, Chen; Deng, Zhicong; Xu, Yaoyang; Zheng, Mengmeng] Xian Univ Posts & Telecommun, Sch Comp Sci & Technol, Shaanxi Key Lab Network Data Anal & Intelligent Pr, Xian 710121, Peoples R China; [Ding, Chen; Deng, Zhicong; Xu, Yaoyang; Zheng, Mengmeng] Xian Univ Posts & Telecommun, Xian Key Lab Big Data & Intelligent Comp, Xian 710121, Peoples R China; [Zhang, Lei; Wei, Wei; Zhang, Yanning] Northwestern Polytech Univ, Sch Comp Sci, Shaanxi Prov Key Lab Speech & Image Informat Proc, Xian 710072, Peoples R China; [Zhang, Lei; Wei, Wei; Zhang, Yanning] Northwestern Polytech Univ, Sch Comp Sci, Natl Engn Lab Integrated Aerosp Ground Ocean Big D, Xian 710072, Peoples R China; [Cao, Yu] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Xian 710119, Peoples R China; [Cao, Yu] Chinese Acad Sci, Key Lab Space Precis Measurement Technol, Xian 710119, Peoples R ChinaAffiliations:Xi'an University of Posts & Telecommunications; Xi'an University of Posts & Telecommunications; Northwestern Polytechnical University; Northwestern Polytechnical University; Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Chinese Academy of SciencesPublication Year:2024Volume:62Article Number:5522519DOI Link:http://dx.doi.org/10.1109/TGRS.2024.3407812數(shù)據(jù)庫ID（收錄號）:WOS:001272260000015 -
Record 353 of
Title:Rapid Determination of Positive-Negative Bacterial Infection Based on Micro-Hyperspectral Technology
Author Full Names:Du, Jian; Tao, Chenglong; Qi, Meijie; Hu, Bingliang; Zhang, ZhoufengSource Title:SENSORSLanguage:EnglishDocument Type:ArticleAbstract:To meet the demand for rapid bacterial detection in clinical practice, this study proposed a joint determination model based on spectral database matching combined with a deep learning model for the determination of positive-negative bacterial infection in directly smeared urine samples. Based on a dataset of 8124 urine samples, a standard hyperspectral database of common bacteria and impurities was established. This database, combined with an automated single-target extraction, was used to perform spectral matching for single bacterial targets in directly smeared data. To address the multi-scale features and the need for the rapid analysis of directly smeared data, a multi-scale buffered convolutional neural network, MBNet, was introduced, which included three convolutional combination units and four buffer units to extract the spectral features of directly smeared data from different dimensions. The focus was on studying the differences in spectral features between positive and negative bacterial infection, as well as the temporal correlation between positive-negative determination and short-term cultivation. The experimental results demonstrate that the joint determination model achieved an accuracy of 97.29%, a Positive Predictive Value (PPV) of 97.17%, and a Negative Predictive Value (NPV) of 97.60% in the directly smeared urine dataset. This result outperformed the single MBNet model, indicating the effectiveness of the multi-scale buffered architecture for global and large-scale features of directly smeared data, as well as the high sensitivity of spectral database matching for single bacterial targets. The rapid determination solution of the whole process, which combines directly smeared sample preparation, joint determination model, and software analysis integration, can provide a preliminary report of bacterial infection within 10 min, and it is expected to become a powerful supplement to the existing technologies of rapid bacterial detection.Addresses:[Du, Jian; Tao, Chenglong; Qi, Meijie; Hu, Bingliang; Zhang, Zhoufeng] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Key Lab Spectral Imaging Technol CAS, Xian 710119, Peoples R China; [Du, Jian; Tao, Chenglong; Qi, Meijie; Hu, Bingliang; Zhang, Zhoufeng] Xian Key Lab Biomed Spect, Xian 710119, Peoples R ChinaAffiliations:Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CASPublication Year:2024Volume:24Issue:2Article Number:507DOI Link:http://dx.doi.org/10.3390/s24020507數(shù)據(jù)庫ID（收錄號）:WOS:001150870900001 -
Record 354 of
Title:High Accurate and Efficient 3D Network for Image Reconstruction of Diffractive-Based Computational Spectral Imaging
Author Full Names:Fan, Hao; Li, Chenxi; Xu, Huangrong; Zhao, Lvrong; Zhang, Xuming; Jiang, Heng; Yu, WeixingSource Title:IEEE ACCESSLanguage:EnglishDocument Type:ArticleAbstract:Diffractive optical imaging spectroscopy as a promising miniaturized and high throughput portable spectral imaging technique suffers from the problem of low precision and slow speed, which limits its wide use in various applications. To reconstruct the diffractive spectral image more accurately and fast, a three-dimensional spectrum recovery algorithm is proposed in this paper. The algorithm takes advantage of a neural network for image reconstruction which consists of a U-Net architecture with 3D convolutional layers to improve the processing precision and speed. Numerical experiments are conducted to prove its effectiveness. It is shown that the mean peak signal-to-noise ratio (MPSNR) of the recovered image relative to the original image is improved by 1.8 dB in comparison to other traditional methods. In addition, the obtained mean structural similarity (MSSIM) of 0.91 meets the standard of discrimination to human eyes. Moreover, the algorithm runs in just 0.36 s, which is faster than other traditional methods. 3D convolutional networks play a critical role in performance improvement. Improvements in processing speed and accuracy have greatly benefited the realization and application of diffractive optical imaging spectroscopy. The new algorithm with high accuracy and fast speed has a great potential application in diffraction lens spectroscopy and paves a new way for emerging more portable spectral imaging technique.Addresses:[Fan, Hao; Li, Chenxi; Xu, Huangrong; Zhao, Lvrong; Yu, Weixing] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Key Lab Spectral Imaging Technol, Xian 710119, Peoples R China; [Fan, Hao; Zhao, Lvrong; Yu, Weixing] Univ Chinese Acad Sci, Ctr Mat Sci & Optoelect Engn, Beijing 100049, Peoples R China; [Zhang, Xuming; Jiang, Heng] Hong Kong Polytech Univ, Dept Appl Phys, Hong Kong, Peoples R ChinaAffiliations:Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Chinese Academy of Sciences; University of Chinese Academy of Sciences, CAS; Hong Kong Polytechnic UniversityPublication Year:2024Volume:12Start Page:120720End Page:120728DOI Link:http://dx.doi.org/10.1109/ACCESS.2024.3451560數(shù)據(jù)庫ID（收錄號）:WOS:001311194400001 -
Record 355 of
Title:Optical alignment technology for 1-meter accurate infrared magnetic system telescope
Author Full Names:Fu, Xing; Lei, Yu; Li, Hua; E, Kewei; Wang, Peng; Liu, Junpeng; Shen, Yuliang; Wang, DongguangSource Title:JOURNAL OF ASTRONOMICAL TELESCOPES INSTRUMENTS AND SYSTEMSLanguage:EnglishDocument Type:ArticleKeywords Plus:DEROTATORAbstract:Accurate infrared magnetic system (AIMS) is a ground-based solar telescope with the effective aperture of 1 m. The system has complex optical path and contains multiple aspherical mirrors. Since some mirrors are anisotropic in space, parallel light undergoes complex spatial reflection after passing through the optical pupil. It is also required that part of the optical axis coincides with the mechanical rotation axis. The system is difficult to align. This article proposes two innovative alignment methods. First, a modularized alignment method is presented. Each module is individually assembled with optical reference reserved. System integration can be completed through optical reference of each module. Second, computer-aided alignment technology is adopted to achieve perfect wavefront. By perturbing the secondary mirror (M2), the influence of M2 position on the wavefront is measured and the mathematical relationship is obtained. Based on the measured wavefront data, the least squares method is used to calculate the M2 alignment and multiple adjustments have been made to M2. The final system wavefront has reached RMS = 0.12 lambda@632.8nm. Through observations of stars and sunspots, it has been demonstrated that the optical system has good wavefront quality. The observed sunspot is clear with the penumbral and umbra discernible. The proposed method has been verified and provides an effective alignment solution for complex off-axis telescope with large aperture. (c) 2024 Society of Photo-Optical Instrumentation Engineers (SPIE)Addresses:[Fu, Xing; Lei, Yu; Li, Hua; E, Kewei; Wang, Peng; Liu, Junpeng] Xian Inst Opt & Precis Mech, Xian, Peoples R China; [Lei, Yu] Univ Chinese Acad Sci, Beijing, Peoples R China; [Shen, Yuliang; Wang, Dongguang] Chinese Acad Sci, Natl Astron Observ, Beijing, Peoples R ChinaAffiliations:Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Chinese Academy of Sciences; University of Chinese Academy of Sciences, CAS; Chinese Academy of Sciences; National Astronomical Observatory, CASPublication Year:2024Volume:10Issue:1Article Number:14004DOI Link:http://dx.doi.org/10.1117/1.JATIS.10.1.014004數(shù)據(jù)庫ID（收錄號）:WOS:001294608100011 -
Record 356 of
Title:Mural Anomaly Region Detection Algorithm Based on Hyperspectral Multiscale Residual Attention Network
Author Full Names:Guo, Bolin; Qiu, Shi; Zhang, Pengchang; Tang, XingjiaSource Title:CMC-COMPUTERS MATERIALS & CONTINUALanguage:EnglishDocument Type:ArticleKeywords Plus:LOW-RANK; TENSORAbstract:Mural paintings hold significant historical information and possess substantial artistic and cultural value. However, murals are inevitably damaged by natural environmental factors such as wind and sunlight, as well as by human activities. For this reason, the study of damaged areas is crucial for mural restoration. These damaged regions differ significantly from undamaged areas and can be considered abnormal targets. Traditional manual visual processing lacks strong characterization capabilities and is prone to omissions and false detections. Hyperspectral imaging can reflect the material properties more effectively than visual characterization methods. Thus, this study employs hyperspectral imaging to obtain mural information and proposes a mural anomaly detection algorithm based on a hyperspectral multi-scale residual attention network (HM-MRANet). The innovations of this paper include: (1) Constructing mural painting hyperspectral datasets. (2) Proposing a multi-scale residual spectral-spatial feature extraction module based on a 3D CNN (Convolutional Neural Networks) network to better capture multiscale information and improve performance on small-sample hyperspectral datasets. (3) Proposing the Enhanced Residual Attention Module (ERAM) to address the feature redundancy problem, enhance the network's feature discrimination ability, and further improve abnormal area detection accuracy. The experimental results show that the AUC (Area Under Curve), Specificity, and Accuracy of this paper's algorithm reach 85.42%, 88.84%, and 87.65%, respectively, on this dataset. These results represent improvements of 3.07%, 1.11% and 2.68% compared to the SSRN algorithm, demonstrating the effectiveness of this method for mural anomaly detection.Addresses:[Guo, Bolin; Qiu, Shi; Zhang, Pengchang] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Key Lab Spectral Imaging Technol CAS, Xian 710119, Peoples R China; [Guo, Bolin] Univ Chinese Acad Sci, Sch Optoelect, Beijing 100408, Peoples R China; [Tang, Xingjia] Northwestern Polytech Univ, Inst Culture & Heritage, Xian 710072, Peoples R ChinaAffiliations:Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Chinese Academy of Sciences; University of Chinese Academy of Sciences, CAS; Northwestern Polytechnical UniversityPublication Year:2024Volume:81Issue:1Start Page:1809End Page:1833DOI Link:http://dx.doi.org/10.32604/cmc.2024.056706數(shù)據(jù)庫ID（收錄號）:WOS:001350270600048 -
Record 357 of
Title:Location-Guided Dense Nested Attention Network for Infrared Small Target Detection
Author Full Names:Guo, Huinan; Zhang, Nengshuang; Zhang, Jing; Zhang, Wuxia; Sun, CongyingSource Title:IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSINGLanguage:EnglishDocument Type:ArticleKeywords Plus:MODELAbstract:Infrared small target (IST) detection involves identifying objects that occupy fewer than 81 pixels in a 256 x 256 image. Because the target is small and lacks texture, structure, and shape information on its surface, this task is highly challenging. CNN-based methods can extract rich features of the target. However, overly deep network structures may increase the risk of losing small targets. In addition, pixel-level positional deviations can also reduce the detection accuracy of IST. To address these challenges, we propose the location-guided dense nested attention network for IST detection. The proposed network consists of a pixel attention guided feature extraction module (PAG-FEM), a channel attention guided feature fusion module (CAG-FFM), and a detection module. First, the PAG-FEM utilizes the DNIM dense nested blocks from the DNANet as the backbone, integrating both channel and pixel attention mechanisms. This method focuses on the semantic and positional information of the targets, yielding semantic features that emphasize the positions of small targets. Second, the CAG-FFM employs upsampling and convolution operations to align the feature sizes, while utilizing the channel attention mechanism to obtain effective channel information. Then, these features are fused through stacking, addition, and averaging operations to obtain more discriminative features. Finally, the detection module uses eight-connected neighborhood clustering method to obtain the centroid coordinates of the targets for subsequent detection evaluation. Three datasets are utilized to verify our method, and experimental results show that our method performs better than other advanced methods.Addresses:[Guo, Huinan] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Xian 710121, Peoples R China; [Zhang, Nengshuang; Zhang, Jing; Sun, Congying] Xian Univ Technol, Automat & Informat Engn, Xian 710048, Peoples R China; [Zhang, Wuxia] Xian Univ Posts & Telecommun, Sch Comp Sci & Technol, Shaanxi Key Lab Network Data Anal & Intelligent Pr, Xian 710121, Peoples R ChinaAffiliations:Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Xi'an University of Technology; Xi'an University of Posts & TelecommunicationsPublication Year:2024Volume:17Start Page:18535End Page:18548DOI Link:http://dx.doi.org/10.1109/JSTARS.2024.3472041數(shù)據(jù)庫ID（收錄號）:WOS:001340861900011 -
Record 358 of
Title:CMID: Crossmodal Image Denoising via Pixel-Wise Deep Reinforcement Learning
Author Full Names:Guo, Yi; Gao, Yuanhang; Hu, Bingliang; Qian, Xueming; Liang, DongSource Title:SENSORSLanguage:EnglishDocument Type:ArticleKeywords Plus:SPARSE; NETWORKAbstract:Removing noise from acquired images is a crucial step in various image processing and computer vision tasks. However, the existing methods primarily focus on removing specific noise and ignore the ability to work across modalities, resulting in limited generalization performance. Inspired by the iterative procedure of image processing used by professionals, we propose a pixel-wise crossmodal image-denoising method based on deep reinforcement learning to effectively handle noise across modalities. We proposed a similarity reward to help teach an optimal action sequence to model the step-wise nature of the human processing process explicitly. In addition, We designed an action set capable of handling multiple types of noise to construct the action space, thereby achieving successful crossmodal denoising. Extensive experiments against state-of-the-art methods on publicly available RGB, infrared, and terahertz datasets demonstrate the superiority of our method in crossmodal image denoising.Addresses:[Guo, Yi; Hu, Bingliang] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Xian 710119, Peoples R China; [Guo, Yi; Qian, Xueming] Xi An Jiao Tong Univ, Sch Informat & Commun Engn, Xian 710049, Peoples R China; [Guo, Yi; Hu, Bingliang] Univ Chinese Acad Sci, Beijing 100049, Peoples R China; [Gao, Yuanhang; Liang, Dong] Nanjing Univ Aeronaut & Astronaut, Coll Comp Sci & Technol, Nanjing 211106, Peoples R ChinaAffiliations:Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CAS; Xi'an Jiaotong University; Chinese Academy of Sciences; University of Chinese Academy of Sciences, CAS; Nanjing University of Aeronautics & AstronauticsPublication Year:2024Volume:24Issue:1Article Number:42DOI Link:http://dx.doi.org/10.3390/s24010042數(shù)據(jù)庫ID（收錄號）:WOS:001140597600001 -
Record 359 of
Title:Rapid Solidification of Invar Alloy
Author Full Names:He, Hanxin; Yao, Zhirui; Li, Xuyang; Xu, JunfengSource Title:MATERIALSLanguage:EnglishDocument Type:ArticleAbstract:The Invar alloy has excellent properties, such as a low coefficient of thermal expansion, but there are few reports about the rapid solidification of this alloy. In this study, Invar alloy solidification at different undercooling (Delta T) was investigated via glass melt-flux techniques. The sample with the highest undercooling of Delta T = 231 K (recalescence height 140 K) was obtained. The thermal history curve, microstructure, hardness, grain number, and sample density of the alloy were analyzed. The results show that with the increase in solidification undercooling, the XRD peak of the sample shifted to the left, indicating that the lattice constant increased and the solid solubility increased. As the solidification of undercooling increases, the microstructure changes from large dendrites to small columnar grains and then to fine equiaxed grains. At the same time, the number of grains also increases with the increase in the undercooling. The hardness of the sample increases with increasing undercooling. If Delta T >= 181 K (128 K), the grain number and the hardness do not increase with undercooling.Addresses:[He, Hanxin] Xian Univ Architecture & Technol, Sch Civil Engn, 13 Yanta Rd, Xian 710055, Peoples R China; [Yao, Zhirui; Xu, Junfeng] Xian Technol Univ, Sch Mat & Chem Engn, Xian 710021, Peoples R China; [Li, Xuyang] Xian Inst Opt & Precis Mech, Xian 710119, Peoples R ChinaAffiliations:Xi'an University of Architecture & Technology; Xi'an Technological University; Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CASPublication Year:2024Volume:17Issue:1Article Number:231DOI Link:http://dx.doi.org/10.3390/ma17010231數(shù)據(jù)庫ID（收錄號）:WOS:001140714800001 -
Record 360 of
Title:Hyperspectral Image Based Interpretable Feature Clustering Algorithm
Author Full Names:Kang, Yaming; Ye, Peishun; Bai, Yuxiu; Qiu, ShiSource Title:CMC-COMPUTERS MATERIALS & CONTINUALanguage:EnglishDocument Type:ArticleKeywords Plus:CLASSIFICATION; DIAGNOSISAbstract:Hyperspectral imagery encompasses spectral and spatial dimensions, reflecting the material properties of objects. Its application proves crucial in search and rescue, concealed target identification, and crop growth analysis. Clustering is an important method of hyperspectral analysis. The vast data volume of hyperspectral imagery, coupled with redundant information, poses significant challenges in swiftly and accurately extracting features for subsequent analysis. The current hyperspectral feature clustering methods, which are mostly studied from space or spectrum, do not have strong interpretability, resulting in poor comprehensibility of the algorithm. So, this research introduces a feature clustering algorithm for hyperspectral imagery from an interpretability perspective. It commences with a simulated perception process, proposing an interpretable band selection algorithm to reduce data dimensions. Following this, a multi-dimensional clustering algorithm, rooted in fuzzy and kernel clustering, is developed to highlight intra-class similarities and inter-class differences. An optimized P system is then introduced to enhance computational efficiency. This system coordinates all cells within a mapping space to compute optimal cluster centers, facilitating parallel computation. This approach diminishes sensitivity to initial cluster centers and augments global search capabilities, thus preventing entrapment in local minima and enhancing clustering performance. Experiments conducted on 300 datasets, comprising both real and simulated data. The results show that the average accuracy (ACC) of the proposed algorithm is 0.86 and the combination measure (CM) is 0.81.Addresses:[Kang, Yaming; Ye, Peishun; Bai, Yuxiu] Yulin Univ, Sch Informat Engn, Yulin 719000, Peoples R China; [Qiu, Shi] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Key Lab Spectral Imaging Technol CAS, Xian 710119, Peoples R ChinaAffiliations:Yulin University; Chinese Academy of Sciences; Xi'an Institute of Optics & Precision Mechanics, CASPublication Year:2024Volume:79Issue:2Start Page:2151End Page:2168DOI Link:http://dx.doi.org/10.32604/cmc.2024.049360數(shù)據(jù)庫ID（收錄號）:WOS:001240838500018