检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Wang, Yunhe Huang, Mingqiang Han, Kai Chen, Hanting Zhang, Wei Xu, Chunjing Tao, Dacheng Noah’s Ark Lab Huawei Technologies China Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China School of Computer Science Faculty of Engineering University of Sydney Australia

Convolutional neural networks (CNN) have been widely used for boosting the performance of many machine intelligence tasks. However, the CNN models are usually computationally intensive and energy consuming, since they are often designed with numerous multiply-operations and considerable parameters for the accuracy reason. Thus, it is difficult to directly apply them in the resource-constrained environments such as’Internet of Things’ (IoT) devices and smart phones. To reduce the computational complexity and energy burden, here we present a novel minimalist hardware architecture using adder convolutional neural network (AdderNet), in which the original convolution is replaced by adder kernel using only additions. To maximally excavate the potential energy consumption, we explore the low-bit quantization algorithm for AdderNet with shared-scaling-factor method, and we design both specific and general-purpose hardware accelerators for AdderNet. Experimental results show that the adder kernel with int8/int16 quantization also exhibits high performance, meanwhile consuming much less resources (theoretically ∼81% off). In addition, we deploy the quantized AdderNet on FPGA (Field Programmable Gate Array) platform. The whole AdderNet can practically achieve 16% enhancement in speed, 67.6%-71.4% decrease in logic resource utilization and 47.85%-77.9% decrease in power consumption compared to CNN under the same circuit architecture. With a comprehensive comparison on the performance, power consumption, hardware resource consumption and network generalization capability, we conclude the AdderNet is able to surpass all the other competitors including the classical CNN, novel memristor-network, XNOR-Net and the shift-kernel based network, indicating its great potential in future high performance and energy-efficient artificial intelligence applications. Copyright © 2021, The Authors. All rights reserved.

关键词： Electric power utilization

来源：评论

学校读者我要写书评

暂无评论

Janus faces of dipolar sources in directional near-field coupling with an oriented misalignment

引用

Physical Review A 2024年第3期109卷 033505-033505页

作者： Chenxu Bian Yuhan Zhong Xuhuinan Chen Tony Low Hongsheng Chen Baile Zhang Xiao Lin State Key Laboratory of Extreme Photonics and Instrumentation ZJU-Hangzhou Global Scientific and Technological Innovation Center College of Information Science and Electronic Engineering Zhejiang University Hangzhou 310027 China and International Joint Innovation Center The Electromagnetics Academy at Zhejiang University Zhejiang University Haining 314400 China Department of Electrical and Computer Engineering University of Minnesota Minneapolis Minnesota 55455 USA Key Lab of Advanced Micro/Nano Electronic Devices & Smart Systems of Zhejiang Jinhua Institute of Zhejiang University Zhejiang University Jinhua 321099 China and Shaoxing Institute of Zhejiang University Zhejiang University Shaoxing 312000 China Division of Physics and Applied Physics School of Physical and Mathematical Sciences and Centre for Disruptive Photonic Technologies Nanyang Technological University Singapore 637371 Singapore

Directional near-field coupling can enable many applications, such as directional routing, on-chip information processing, and chiral quantum optics. This directional near-field coupling is generally face dependent, since the induced near-field radiation pattern is dependent on which face (e.g., the upper or lower face) of the dipolar source is facing toward the outcoupler. In other words, the dipolar sources in the directional near-field coupling intrinsically have two faces (e.g., Janus faces), whose corresponding near-field coupling strength and coupling directionality might be distinct. Here, we outline a general physical framework to control these Janus faces in the directional near-field coupling and find that they can have the feature of an oriented misalignment. To be specific, the upper and lower radiation patterns of excited surface waves can have the same shape but are misaligned with a twist angle. For example, the twist angle can be any acute angle if the source is composed of a Huygens dipole and an electric dipole, and it can be any obtuse angle if the source is constructed by a circular electric dipole and a magnetic dipole.

关键词： Classical optics Light-matter interaction Metamaterials Near-field optics Photonics Plasmonics

来源：评论

学校读者我要写书评

暂无评论

Probabilistic margins for instance reweighting in adversarial training 21

Probabilistic margins for instance reweighting in adversaria...

引用

Proceedings of the 35th International Conference on Neural Information Processing Systems

作者： Qizhou Wang Feng Liu Bo Han Tongliang Liu Chen Gong Gang Niu Mingyuan Zhou Masashi Sugiyama Department of Computer Science Hong Kong Baptist University DeSI Lab Australian Artificial Intelligence Institute University of Technology Sydney TML Lab School of Computer Science Faculty of Engineering The University of Sydney PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of MoE and Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology RIKEN Center for Advanced Intelligence Project (AIP) McCombs School of Business The University of Texas at Austin RIKEN Center for Advanced Intelligence Project (AIP) and Graduate School of Frontier Sciences The University of Tokyo

ISBN: (纸本)9781713845393

Reweighting adversarial data during training has been recently shown to improve adversarial robustness, where data closer to the current decision boundaries are regarded as more critical and given larger weights. However, existing methods measuring the closeness are not very reliable: they are discrete and can take only a few values, and they are path-dependent, i.e., they may change given the same start and end points with different attack paths. In this paper, we propose three types of probabilistic margin (PM), which are continuous and path-independent, for measuring the aforementioned closeness and reweighting adversarial data. Specifically, a PM is defined as the difference between two estimated class-posterior probabilities, e.g., such a probability of the true label minus the probability of the most confusing label given some natural data. Though different PMs capture different geometric properties, all three PMs share a negative correlation with the vulnerability of data: data with larger/smaller PMs are safer/riskier and should have smaller/larger weights. Experiments demonstrated that PMs are reliable and PM-based reweighting methods outperformed state-of-the-art counterparts.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Novel view synthesis from a single image via unsupervised learning

arXiv

引用

arXiv 2021年

作者： Liu, Bingzheng Lei, Jianjun Peng, Bo Yu, Chuanbo Li, Wanqing Ling, Nam School of Electrical and Information Engineering Tianjin University Tianjin300072 China Advanced Multimedia Research Lab University of Wollongong Wollongong Australia Department of Computer Science and Engineering Santa Clara University Santa ClaraCA95053 United States

View synthesis aims to generate novel views from one or more given source views. Although existing methods have achieved promising performance, they usually require paired views of different poses to learn a pixel transformation. This paper proposes an unsupervised network to learn such a pixel transformation from a single source viewpoint. In particular, the network consists of a token transformation module (TTM) that facilities the transformation of the features extracted from a source viewpoint image into an intrinsic representation with respect to a pre-defined reference pose and a view generation module (VGM) that synthesizes an arbitrary view from the representation. The learned transformation allows us to synthesize a novel view from any single source viewpoint image of unknown pose. Experiments on the widely used view synthesis datasets have demonstrated that the proposed network is able to produce comparable results to the state-of-the-art methods despite the fact that learning is unsupervised and only a single source viewpoint image is required for generating a novel view. The code will be available soon. Copyright © 2021, The Authors. All rights reserved.

关键词： Three dimensional displays

来源：评论

学校读者我要写书评

暂无评论

Intelligent Multi-Modal Sensing-Communication Integration: Synesthesia of Machines

arXiv

引用

arXiv 2023年

作者： Cheng, Xiang Zhang, Haotian Zhang, Jianan Gao, Shijian Li, Sijiang Huang, Ziwei Bai, Lu Yang, Zonghui Zheng, Xinhu Yang, Liuqing The State Key Laboratory of Advanced Optical Communication Systems and Networks School of Electronics Peking University Beijing100871 China The Samsung Semiconductor Samsung SoC Research and Development Lab San DiegoCA92121 United States Shandong University Jinan250100 China Guangzhou510000 China Guangzhou510000 China The Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology Hong Kong

In the era of sixth-generation (6G) wireless communications, integrated sensing and communications (ISAC) is recognized as a promising solution to upgrade the physical system by endowing wireless communications with sensing capability. Existing ISAC is mainly oriented to static scenarios with radio-frequency (RF) sensors being the primary participants, thus lacking a comprehensive environment feature characterization and facing a severe performance bottleneck in dynamic environments. To date, extensive surveys on ISAC have been conducted but are limited to summarizing RF-based radar sensing. Currently, some research efforts have been devoted to exploring multi-modal sensing-communication integration but still lack a comprehensive review. To fill the gap, we embark on an initial endeavor with the goal of establishing a unified framework of intelligent multi-modal sensing-communication integration by generalizing the concept of ISAC and providing a comprehensive review under this framework. Inspired by the human synesthesia, the so-termed Synesthesia of Machines (SoM) gives the clearest cognition of such an intelligent integration and details its paradigm for the first time. We commence by justifying the necessity and potential of the new paradigm. Subsequently, we offer a rigorous definition of SoM and zoom into the detailed paradigm, which is summarized as three operational modes realizing the integration. To facilitate SoM research, we overview the prerequisite of SoM research, that is, mixed multi-modal (MMM) datasets, and introduce our work. Built upon the MMM datasets, we introduce the mapping relationships between multi-modal sensing and communications, and discuss how channel modeling can be customized to support the exploration of such relationships. Afterward, aiming at giving a comprehensive survey on the current research status of multi-modal sensing-communication integration, we cover the technological review on SoM-enhance-based and SoM-concert-based app

关键词： Beamforming

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study of Control Methods for Grid Side Converters in PMSG-Based Wind Energy Conversion Systems

A Comparative Study of Control Methods for Grid Side Convert...

引用

IEEE International Symposium on Industrial Electronics (ISIE)

作者： Mohammad Sadegh Eslahi Sadegh Vaez-Zadeh Alireza Jabbarnejad Advanced Motion System Research Lab School of Electrical and Computer Engineering College of Engineering University of Tehran Tehran Iran Advanced Motion System Research Lab and the Center of Excellence on Applied Electromagnetic Systems School of Electrical and Computer Engineering College of Engineering University of Tehran Tehran Iran

ISBN: (数字)9781728156354

ISBN: (纸本)9781728156361

Wind energy conversion systems are interfaced with the power grids through power electronic converters. The main duty of grid side converters is to deliver the power to the grid, while fulfilling the grid code requirements. In this paper, the grid codes for the integration of wind power to the grids are presented first. Then the principles of three main control methods for the grid side converters, i.e. vector control, direct power control, and combined control, for permanent magnet synchronous generators are reviewed. In addition, their capabilities and weaknesses in meeting grid codes are investigated. Finally, a comparison of the control performances in fulfilling grid code requirements are presented by extensive simulation results.

关键词： Reactive power Wind turbines Control systems Couplings Wind power generation Power quality Wind energy

来源：评论

学校读者我要写书评

暂无评论

Dynamic Management of Traffic Signals through Social IoT 3

Dynamic Management of Traffic Signals through Social IoT

引用

3rd International Conference on Computing and Network Communications, CoCoNet 2019

作者： Roopa, M.S. Ayesha Siddiq, S. Buyya, Rajkumar Venugopal, K.R. Iyengar, S.S. Patnaik, L.M. IoT Lab University Visvesvaraya College of Engineering Bangalore University Bengaluru India Lab School of Computing and Information Systems University of Melbourne Australia Bangalore University Bengaluru India Department of Computer Science and Engineering Florida International University United States National Institute of Advanced Studies Indian Institute of Science Campus Bengaluru India

Traffic congestion is a major threat to transportation sector in every urban city around the world. This causes many adverse effects like, heavy fuel consumption, increased waiting time, pollution, etc. and pose an eminent challenge to the movement of emergency vehicles. To achieve better driving we proceed towards a trending research field called Social Internet of Vehicles (SIoV). A social network paradigm that permits the establishment of social relationships among every vehicle in the network or with any road infrastructure can be radically helpful. This holds as the aim of SIoV, to be beneficial for the drivers, in improving the road safety, avoiding mishaps, and have a friendly-driving environment. In this paper, we propose a Dynamic congestion control with Throughput Maximization scheme based on Social Aspect (D-TMSA) utilizing the social, behavioral and preference-based relationships. Our proposed scheme along with the various social relationship types allocates green signal to maximize the traffic flow passing through an intersection. Simulation results show that the D-TMSA outperforms the existing work by achieving high throughput, lowering the total traveling time and reducing the average waiting time to better the flow of traffic based on their social attributes with each other. © 2020 The Authors. Published by Elsevier B.V.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

Wander 22

Wander

引用

Proceedings of the SIGGRAPH Asia 2022 Art Gallery

作者： Yuqian Sun Chang Hee Lee Chenhang Cheng Ali Asadipour Ying Xu Yihua Li Computer Science Research Centre Royal College of Art London United Kingdom Affective Systems and Cognition Lab Industrial Design Department College of Engineering Korea Advanced Institute of Science and Technology Daejeon Republic of Korea Unaffiliated Shanghai China Department of Industrial Design School of Art and Design Wuhan University of Technology Wuhan China Department of Product Design College of Fashion and Design Donghua University Shanghai China

This is a public art project centred on an AI chatbot called Wander. She can generate interactive fiction based on realworld locations' knowledge graphs. Through public participation, the stories are visualised on...

ISBN: (纸本)9781450394710

关键词：

来源：评论

学校读者我要写书评

暂无评论

Contrastive Bayesian Analysis for Deep Metric Learning

arXiv

引用

arXiv 2022年

作者： Kan, Shichao He, Zhiquan Cen, Yigang Li, Yang Mladenovic, Vladimir He, Zhihai School of Computer Science and Engineering Central South University Hunan Changsha410083 China Institute of Information Science School of Computer and Information Technology Beijing Jiaotong University Beijing100044 China Beijing Key Laboratory of Advanced Information Science and Network Technology Beijing100044 China Guangdong Multimedia Information Service Engineering Technology Research Center Shenzhen University 518060 China Department of Electrical Engineering and Computer Science University of Missouri ColumbiaMO65211 United States Faculty of Technical Sciences University of Kragujevac Cacak Serbia Department of Electrical and Electronic Engineering Southern University of Science and Technology Shenzhen China Pengcheng Lab Shenzhen518066 China

Recent methods for deep metric learning have been focusing on designing different contrastive loss functions between positive and negative pairs of samples so that the learned feature embedding is able to pull positive samples of the same class closer and push negative samples from different classes away from each other. In this work, we recognize that there is a significant semantic gap between features at the intermediate feature layer and class labels at the final output layer. To bridge this gap, we develop a contrastive Bayesian analysis to characterize and model the posterior probabilities of image labels conditioned by their features similarity in a contrastive learning setting. This contrastive Bayesian analysis leads to a new loss function for deep metric learning. To improve the generalization capability of the proposed method onto new classes, we further extend the contrastive Bayesian loss with a metric variance constraint. Our experimental results and ablation studies demonstrate that the proposed contrastive Bayesian metric learning method significantly improves the performance of deep metric learning in both supervised and pseudo-supervised scenarios, outperforming existing methods by a large margin. © 2022, CC0.

关键词： Distance education

来源：评论

学校读者我要写书评

暂无评论

In-silico study of lanthanide-based nanoparticles for dual-modal photoacoustic and MRI theranostics

引用

Scientific reports 2025年第1期15卷 18818页

作者： Farid Alidoust Aghdam Ali Rostami Photonics and Nanocrystals Research Lab (PNRL) Faculty of Electrical and Computer Engineering University of Tabriz Tabriz Iran. Photonics and Nanocrystals Research Lab (PNRL) Faculty of Electrical and Computer Engineering University of Tabriz Tabriz Iran. rostami@tabrizu.ac.ir. SP-EPT Lab. ASEPE Company Industrial Park of Advanced Technologies Tabriz Iran. rostami@tabrizu.ac.ir.

This study presents a comprehensive theoretical assessment of lanthanide-based core-shell nanoparticles for theranostic applications in the near-infrared region. The complex refractive indices of fifteen lanthanide elements were calculated using density functional theory and the hybrid HSE06 method, which provided optical data in the spectral range of 100-2500 nm. Mie's theoretical calculations for various biocompatible coatings (SiO, PEG, TiO) indicated that TiO coatings exhibit better absorption efficiency in the NIR-I window (750-900 nm) (Q up to 9.4×10). The optimal core size for TiO-coated nanoparticles was determined to be between 90 and 110 nm, and the core-shell ratio ranged from 0.54 to 0.63, which provides the highest adsorption efficiency. Modeling biological heat transfer using the Pennes model with the finite element method revealed that the thermal reactions of tissues vary significantly for each tissue type. The thermal threshold time followed a power relation with . The fastest thermal response was observed in breast tissue (threshold time: 2.40-9.60 seconds), while the slowest response was noted in liver tissue (threshold time: 19.20-4.80 seconds). These findings offer key parameters for optimizing lanthanide-based therapeutic imaging platforms and provide a theoretical framework for predicting their performance in biological environments.

关键词： Bio-heat transfer Core-shell nanostructures Density functional theory Lanthanide nanoparticles Mie theory Near-infrared window Photothermal therapy

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：