检索结果-内蒙古大学图书馆

Stacked auto encoder based Deep Reinforcement Learning for online resource scheduling in large-scale MEC networks

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Jiang, Feibo Wang, Kezhi Dong, Li Pan, Cunhua Yang, Kun Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing Hunan Normal University Changsha China Department of Computer and Information Sciences Northumbria University United Kingdom Key Laboratory of Hunan Province for New Retail Virtual Reality Technology Hunan University of Commerce Changsha China School of Electronic Engineering and Computer Science Queen Mary University of London LondonE1 4NS United Kingdom School of Computer Sciences and Electrical Engineering University of Essex ColchesterCO4 3SQ United Kingdom University of Electronic Science and Technology of China Chengdu China

An online resource scheduling framework is proposed for minimizing the sum of weighted task latency for all the mobile users, by optimizing offloading decision, transmission power, and resource allocation in the mobile edge computing (MEC) system. Towards this end, a deep reinforcement learning (DRL) method is proposed to obtain an online resource scheduling policy. Firstly, a related and regularized stacked auto encoder (2r-SAE) with unsupervised learning is proposed to perform data compression and representation for high dimensional channel quality information (CQI) data, which can reduce the state space for DRL. Secondly, we present an adaptive simulated annealing based approach (ASA) as the action search method of DRL, in which an adaptive h-mutation is used to guide the search direction and an adaptive iteration is proposed to enhance the search efficiency during the DRL process. Thirdly, a preserved and prioritized experience replay (2p-ER) is introduced to assist the DRL to train the policy network and find the optimal offloading policy. Numerical results are provided to demonstrate that the proposed algorithm can achieve near-optimal performance while significantly decreasing the computational time compared with existing benchmarks. It also shows that the proposed framework is suitable for resource scheduling problem in large-scale MEC networks, especially in the dynamic environment. Copyright © 2020, The Authors. All rights reserved.

关键词： Deep learning

Designing and Training of A Dual CNN for Image Denoising

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zuo, Wangmeng Zhang, David Lin, Chia-Wen Xu, Yong Tian, Chunwei Du, Bo Bio-Computing Research Center Harbin Institute of Technology Shenzhen China Shenzhen Key Laboratory of Visual Object Detection and Recognition ShenzhenGuangdong518055 China Peng Cheng Laboratory Shenzhen518055 China School of Computer Science and Technology Harbin Institute of Technology 150001 HarbinHeilongjiang China Peng Cheng Laboratory Shenzhen518055 China School of Computer Science Wuhan University 430072 WuhanHubei China Department of Electrical Engineering Institute of Communications Engineering National Tsing Hua University Hsinchu Taiwan 518172 ShenzhenGuangdong China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen China

Deep convolutional neural networks (CNNs) for image denoising have recently attracted increasing research interest. However, plain networks cannot recover fine details for a complex task, such as real noisy images. In this paper, we propose a Dual denoising Network (DudeNet) to recover a clean image. Specifically, DudeNet consists of four modules: a feature extraction block, an enhancement block, a compression block, and a reconstruction block. The feature extraction block with a sparse mechanism extracts global and local features via two sub-networks. The enhancement block gathers and fuses the global and local features to provide complementary information for the latter network. The compression block refines the extracted information and compresses the network. Finally, the reconstruction block is utilized to reconstruct a denoised image. The DudeNet has the following advantages: (1) The dual networks with a parse mechanism can extract complementary features to enhance the generalized ability of denoiser. (2) Fusing global and local features can extract salient features to recover fine details for complex noisy images. (3) A Small-size filter is used to reduce the complexity of denoiser. Extensive experiments demonstrate the superiority of DudeNet over existing current state-of-the-art denoising methods. Copyright © 2020, The Authors. All rights reserved.

关键词： Recovery

Detecting out-of-distribution samples via variational auto-encoder with reliable uncertainty estimation

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Ran, Xuming Xu, Mingkun Mei, Lingrui Xu, Qi Liu, Quanying Shenzhen Key Laboratory of Smart Healthcare Engineering Department of Biomedical Engineering Southern University of Science and Technology Shenzhen518055 China Center for Brain Inspired Computing Research Department of Precision Instrument Tsinghua University Beijing100084 China China Automotive Engineering Research Institute Chongqing401122 China School of Artifical Intelligence Electronic and Electrical Engineering School of Artifical Intelligence Dalian University of Technology Dalian116024 China College of Mathematics and Statistics Chongqing Jiaotong University Chongqing400074 China College of Computer Science and Technology Zhejiang University Hangzhou310027 China

Variational autoencoders (VAEs) are influential generative models with rich representation capabilities from the deep neural network architecture and Bayesian method. However, VAE models have a weakness that assign a higher likelihood to out-of-distribution (OOD) inputs than in-distribution (ID) inputs. To address this problem, a reliable uncertainty estimation is considered to be critical for in-depth understanding of OOD inputs. In this study, we propose an improved noise contrastive prior (INCP) to be able to integrate into the encoder of VAEs, called INCPVAE. INCP is scalable, trainable and compatible with VAEs, and it also adopts the merits from the INCP for uncertainty estimation. Experiments on various datasets demonstrate that compared to the standard VAEs, our model is superior in uncertainty estimation for the OOD data and is robust in anomaly detection tasks. The INCPVAE model obtains reliable uncertainty estimation for OOD inputs and solves the OOD problem in VAE models. © 2020, CC BY.

关键词： Bayesian networks

Magnetic anisotropy reversal driven by structural symmetry-breaking in monolayer α-RuCl3

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yang, Bowen Goh, Yin Min Sung, Suk Hyun Ye, Gaihua Biswas, Sananda Kaib, David A.S. Dhakal, Ramesh Yan, Shaohua Li, Chenghe Jiang, Shengwei Chen, Fangchu Lei, Hechang He, Rui Valentí, Roser Winter, Stephen M. Hovden, Robert Tsen, Adam W. Institute for Quantum Computing University of Waterloo WaterlooONN2L 3G1 Canada Department of Physics and Astronomy University of Waterloo WaterlooONN2L 3G1 Canada Department of Physics University of Michigan Ann ArborMI48109 United States Department of Materials Science and Engineering University of Michigan Ann ArborMI48109 United States Department of Electrical and Computer Engineering Texas Tech University LubbockTX79409 United States Institut für Theoretische Physik Goethe-Universität Frankfurt Max-von-Laue-Strasse 1 Frankfurt am Main60438 Germany Department of Physics Center for Functional Materials Wake Forest University Winston-SalemNC27109 United States Department of Physics Beijing Key Laboratory of Opto-electronic Functional Materials & Micro-nano Devices Renmin University of China Beijing100872 China Department of Physics Cornell University IthacaNY United States School of Physics and Astronomy Shanghai Jiao Tong University Shanghai China Applied Physics Program University of Michigan Ann ArborMI48109 United States Department of Chemistry University of Waterloo WaterlooONN2L 3G1 Canada

The Kitaev model is a celebrated spin-½ model on the two-dimensional (2D) honeycomb lattice with bond-dependent Ising interactions1, which features a highly entangled quantum spin liquid (QSL) ground state, fractionalized Majorana excitations, and a series of magnetic-field-induced quantum phase transitions1-7. The search for materials realizing the Kitaev model has been an ongoing challenge for over a decade and may potentially lead to applications in fault-tolerant topological quantum computing8. Yet, the unavoidable presence of non-Kitaev interactions (Heisenberg, off-diagonal, next-nearest neighbor, etc.) almost always drives the ground state away from the QSL phase, and a careful tuning of the exchange parameters is needed9-12. α-RuCl3 is a layered van der Waals material that is a particularly promising candidate to realize Kitaev physics13-15. Although the ground state is zigzag (ZZ) antiferromagnetic (AFM), this ordering can be suppressed with the application of an ~ 6-8T in-plane magnetic field. The presence of a half-integer thermal quantum Hall effect has been reported in this intermediate phase at low temperature16,17, while an unusual continuum of magnetic excitations can be seen even without magnetic field that persists far above the Néel temperature (TN ~ 7-8K)18,19. Both observations hint at α-RuCl3 being in proximity to a QSL, making it a current subject of intense scrutiny. Yet from a theoretical point of view, a QSL induced by in-plane field generally cannot be accounted for, as most calculations for α-RuCl3 show Kitaev phases more broadly emerging from an out-of-plane field6,20-23. Due to the strong "easy-plane" magnetic anisotropy of bulk crystals, however, prohibitively high fields above 30T are required to access such states24-26. The strong coupling between the spin, charge, and lattice degrees of freedom in α-RuCl3 presents an exciting opportunity to tune its magnetic interactions via external perturbations. Although pressure27-29, strain30,

关键词： Ground state

N superlattice multilayer: A real optical mode ferromagnetic resonance thick-film

学校读者我要写书评

暂无评论

Applied Physics Letters 2020年第15期116卷

作者： Du, Honglei Xue, Qian Li, Shandong Miao, Guo-Xing Chu, Xian-Ming Dai, Youyong Yu, Meijie Zhao, Guoxia Yan, Shishen School of Physics State Key Laboratory of Crystal Materials Shandong University Jinan250100 China College of Physics Center for Marine Observation and Communication Qingdao University Qingdao266071 China Institute for Quantum Computing Department of Electrical and Computer Engineering University of Waterloo WaterlooN2L3G1 Canada Department of Cardiology Affiliated Hospital Qingdao University Qingdao266071 China School of Materials Science and Engineering Shandong University Jinan250100 China Spintronics Institute University of Jinan Jinan250022 China

A pure optical mode (OM) resonance with ultrahigh resonance frequency (f r O) and permeability μ at zero bias field was observed in 50-nm FeCoB/Ru/FeCoB sandwich trilayers due to the presence of strong antiferromagnetic interlayer exchange coupling. However, the necessary thickness for creating practical microwave devices cannot be readily achieved by increasing the FeCoB layer thickness or stacking FeCoB/Ru units due to dramatically deteriorated high-frequency performance. In this study, several 50-nm FeCoB/Ru/FeCoB OM units were stacked to form a superlattice structure, with 10-nm ZnO insulator spacers separating the units. It is interesting that the superlattice multilayers not only retain excellent high-frequency OM resonance performance comparable to the single OM unit but also increase the effective magnetic film thickness by a factor of 5. This can be attributed to the positive superposition of the almost identical OM units as well as the effective decoupling between them by thick ZnO spacers. This study provides a promising way to fabricate thicker films while still maintaining excellent OM resonance performance. © 2020 Author(s).

关键词： Zinc oxide

Author Correction: The high-dimensional space of human diseases built from diagnosis records and mapped to genetic loci

学校读者我要写书评

暂无评论

Nature computational science 2023年第7期3卷 658页

作者： Gengjie Jia Yu Li Xue Zhong Kanix Wang Milton Pividori Rabab Alomairy Aniello Esposito Hatem Ltaief Chikashi Terao Masato Akiyama Koichi Matsuda David E Keyes Hae Kyung Im Takashi Gojobori Yoichiro Kamatani Michiaki Kubo Nancy J Cox James Evans Xin Gao Andrey Rzhetsky Shenzhen Branch Guangdong Laboratory of Lingnan Modern Agriculture Genome Analysis Laboratory of the Ministry of Agriculture and Rural Affairs Agricultural Genomics Institute at Shenzhen Chinese Academy of Agricultural Sciences Shenzhen China. jiagengjie@***. Computational Bioscience Research Center King Abdullah University of Science and Technology Thuwal Saudi Arabia. Computer Science Program Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal Saudi Arabia. Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong People's Republic of China. Department of Medicine and Vanderbilt Genetics Institute Vanderbilt University Medical Center Nashville TN US. Department of Medicine Institute of Genomics and Systems Biology Committee on Genomics Genetics and Systems Biology University of Chicago Chicago IL US. Department of Operations Business Analytics and Information Systems University of Cincinnati Cincinnati OH US. Department of Systems Pharmacology and Translational Therapeutics Perelman School of Medicine University of Pennsylvania Philadelphia PA US. Extreme Computing Research Center Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal Saudi Arabia. College of Computer Science and Engineering University of Jeddah Jeddah Saudi Arabia. HPE HPC/AI EMEA Research Laboratory Bristol UK. RIKEN Center for Integrative Medical Sciences Yokohama Japan. Clinical Research Center Shizuoka General Hospital Shizuoka Japan. Department of Applied Genetics The School of Pharmaceutical Sciences University of Shizuoka Shizuoka Japan. Department of Ophthalmology Graduate School of Medical Sciences Kyushu University Fukuoka Japan. Department of Computational Biology and Medical Sciences Graduate School of Frontier Sciences The University of Tokyo Tokyo Japan. Biological and Environmenta

Monitoring macroplastics in aquatic and terrestrial ecosystems: Expert survey reveals visual and drone-based census as most effective techniques

学校读者我要写书评

暂无评论

science of the Total Environment 2024年 955卷 176528页

作者： Gallitelli, L. Girard, P. Andriolo, U. Liro, M. Suaria, G. Martin, C. Lusher, A.L. Hancke, K. Blettler, M.C.M. Garcia-Garin, O. Napper, I.E. Corbari, L. Cózar, A. Morales-Caselles, C. González-Fernández, D. Gasperi, J. Giarrizzo, T. Cesarini, G. De, K. Constant, M. Koutalakis, P. Gonçalves, G. Sharma, P. Gundogdu, S. Kumar, R. Garello, N.A. Camargo, A.L.G. Topouzelis, K. Galgani, F. Royer, S.J. Zaimes, G.N. Rotta, F. Lavender, S. Nava, V. Castro-Jiménez, J. Mani, T. Crosti, R. Azevedo-Santos, V.M. Bessa, F. Tramoy, R. Costa, M.F. Corbau, C. Montanari, A. Battisti, C. Scalici, M. Department of Sciences University Roma Tre Viale Guglielmo Marconi 446 Rome00146 Italy Biosciences Institute Federal University of Mato Grosso MT Cuiabá78060-900 Brazil INESC Coimbra Department of Electrical and Computer Engineering Polo 2 Coimbra3030-290 Portugal Institute of Nature Conservation Polish Academy of Sciences al. Adama Mickiewicza 33 Kraków31–120 Poland Istituto di Scienze Marine - Consiglio Nazionale delle Ricerche CNR-ISMAR Pozzuolo di Lerici La Spezia Italy Thuwal Saudi Arabia Oslo Norway Ciudad Universitaria Santa Fe3000 Argentina Faculty of Biology Universitat de Barcelona Barcelona08028 Spain International Marine Litter Research Unit University of Plymouth Plymouth United Kingdom School of Biological and Marine Sciences University of Plymouth Plymouth United Kingdom Dipartimento di Ingegneria Università degli Studi di Palermo Palermo Italy Department of Biology University Marine Research Institute INMAR University of Cádiz and European University of the Seas SEA-EU Puerto Real Spain Univ Gustave Eiffel GERS-EE Campus Nantes France Fortaleza Brazil Corso Tonolli 50 Verbania Pallanza28922 Italy Biological Oceanography Division CSIR- National Institute of Oceanography Dona Paula Goa403004 India Univ. Lille Institut Mines-Télécom Univ. Artois Junia ULR 4515 - LGCgE Laboratoire de Génie Civil et géo-Environnement LilleF-59000 France Department of Forestry and Natural Environment Science International Hellenic University University Campus in Drama Drama66100 Greece University of Coimbra Department of Mathematics Coimbra Portugal Department of Agricultural Engineering and Technology School of Engineering and Technology Nagaland University Nagaland Dimapur India Cukurova University Department of Basic Science Adana Turkey Department of Biosystems Engineering Auburn University AuburnAL36849 United States Cuiabá Brazil Department of Marine Sciences University of Aegean Greece ECHOS D'OCEANS Corse

Anthropogenic litter, such as plastic, is investigated by the global scientific community from various fields employing diverse techniques. The goal is to assess and finally mitigate the pollutants' impacts on the natural environment. Plastic litter can accumulate in different matrices of aquatic and terrestrial ecosystems, impacting both biota and ecosystem functioning. Detection and quantification of macroplastics, and other litter, can be realized by jointly using visual census and remote sensing techniques. The primary objective of this research was to identify the most effective approach for monitoring macroplastic litter in riverine and marine environments through a comprehensive survey based on the experiences of the scientific community. Researchers involved in plastic pollution evaluated four litter occurrence and flux investigation methods (visual census, drone-based surveys, satellite imagery, and GPS/GNSS trackers) through a questionnaire. Traditional visual census and drone deployment were deemed as the most popular approaches among the 46 surveyed researchers, while satellite imagery and GPS/GNSS trackers received lower scores due to limited field validation and short performance ranges, respectively. On a scale from 0 to 5, visual census and drone-based surveys obtained 3.5 and 2.0, respectively, whereas satellite imagery and alternative solutions received scores lower than 1.2. Visual and drone censuses were used in high, medium and low-income countries, while satellite census and GPS/GNSS trackers were mostly used in high-income countries. This work provides an overview of the advantages and drawbacks of litter investigation techniques, contributing i) to the global harmonization of macroplastic litter monitoring and ii) providing a starting point for researchers and water managers approaching this topic. This work supports the selection and design of reliable and cost-effective monitoring approaches to mitigate the ambiguity in macroplastic data co

关键词： Macroinvertebrates

Topological Unwinding in an Exciton-Polariton Condensate Array

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lyu, Guitao Minami, Yuki Kim, Na Young Byrnes, Tim Watanabe, Gentaro School of Physics Zhejiang Institute of Modern Physics Zhejiang University Zhejiang Hangzhou310027 China Division of Natural and Applied Sciences Duke Kunshan University Jiangsu Kunshan215300 China Faculty of Engineering Gifu University 1-1 Yanagido Gifu501-1193 Japan Institute for Quantum Computing University of Waterloo 200 University Ave. West WaterlooONN2L 3G1 Canada Department of Electrical and Computer Engineering University of Waterloo 200 University Ave. West WaterlooONN2L 3G1 Canada Waterloo Institute for Nanotechnology University of Waterloo 200 University Ave. West WaterlooONN2L 3G1 Canada New York University Shanghai NYU-ECNU Institute of Physics at NYU Shanghai Shanghai Frontiers Science Center of Artificial Intelligence and Deep Learning 567 West Yangsi Road Shanghai200126 China State Key Laboratory of Precision Spectroscopy School of Physical and Material Sciences East China Normal University Shanghai200062 China NYUAD Research Institute New York University Abu Dhabi United Arab Emirates Department of Physics New York University New YorkNY10003 United States Zhejiang Province Key Laboratory of Quantum Technology and Device Zhejiang University Zhejiang Hangzhou310027 China

The phase distribution in a Bose-Einstein condensate can realize various topological states classified by distinct winding numbers. While states with different winding numbers are topologically protected in the linear Schrödinger equation, when nonlinearities are introduced, violations of the topological protection can occur, leading to unwinding. Exciton-polariton condensates constitute a nonlinear open-dissipative system that is well suited to studying such physics. Here we show that a one-dimensional array of exciton-polariton condensates displays a spontaneous phase unwinding from a πto zero-state. We clarify that this collective mode transition is caused by the combined effect of nonlinearity and topological defects in the condensates. While the mode-switching phenomenon observed in our previous experiment was interpreted as the single-particle mode competition, we offer an alternative explanation in terms the collective phase unwinding and find its evidence by reanalyzing the experimental data. Our results open a route towards active control of the mode switching by manipulating the topological defects in prospective quantum polaritonic devices. Copyright © 2023, The Authors. All rights reserved.

关键词： Winding

RF-Net: An end-to-end image matching network based on receptive field

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Shen, Xuelun Wang, Cheng Li, Xin Yu, Zenglei Li, Jonathan Wen, Chenglu Cheng, Ming He, Zijian Fujian Key Laboratory of Sensing and Computing for Smart City School of Information Science and Engineering Xiamen University China School of Electrical Engineering and Computer Science Louisiana State University United States Department of Geography and Environmental Management University of Waterloo Canada

This paper proposes a new end-to-end trainable matching network based on receptive field, RF-Net, to compute sparse correspondence between images. Building end-toend trainable matching framework is desirable and challenging. The very recent approach, LF-Net, successfully embeds the entire feature extraction pipeline into a jointly trainable pipeline, and produces the state-of-the-art matching results. This paper introduces two modifications to the structure of LF-Net. First, we propose to construct receptive feature maps, which lead to more effective keypoint detection. Second, we introduce a general loss function term, neighbor mask, to facilitate training patch selection. This results in improved stability in descriptor training. We trained RF-Net on the open dataset HPatches, and compared it with other methods on multiple benchmark datasets. Experiments show that RF-Net outperforms existing state-of-the-art methods. Copyright © 2019, The Authors. All rights reserved.

关键词： Pipelines