检索结果-内蒙古大学图书馆

Understanding the Runtime Overheads of Deep Learning Inference on Edge Devices

学校读者我要写书评

暂无评论

Understanding the Runtime Overheads of Deep Learning Inferen...

IEEE International Conference on Big Data and Cloud Computing (BdCloud)

作者： Xiu Ma Guangli Li Lei Liu Huaxiao Liu Xiaobing Feng College of Computer Science and Technology Jilin University China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University China SKL of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences China University of Chinese Academy of Sciences China

With the growing ubiquity of the Internet of Things, in-the-edge inference of deep neural network models has been a major driver for promoting the widespread use of intelligent applications. As model inference characteristics are crucial for optimizing and deploying deep neural networks on hardware platforms, many studies focus on analyzing the performance of neural networks such as latency, accuracy, throughput, and energy consumption. However, few existing works have ever discussed the runtime overheads hidden in neural network inference, despite the overheads are non-negligible for edge applications. The lack of in-depth analysis of the overheads hinders the understanding of how hardware designs and model structures impact on-device inference performance. In this paper, we characterize the runtime overheads of deep learning inference on representative edge devices by leveraging state-of-the-art neural network models, performing a systematical analysis from the perspective of end-to-end performance, hardware platforms, memory bandwidth, and neural network model structures. Based on experimental results, the crucial insights are offered to facilitate the design and configure of resource-efficient networks and pick appropriate models on the specific platform, which provides a comprehensive view of runtime overheads of in-the- edge neural network inference for architects and developers.

关键词： Deep learning Performance evaluation Analytical models Runtime Roads Neural networks Throughput

Numerical Simulation of Self-balanced Warm-Air Heating in Boiler-Room of Thermal Power Plant in Winter 11th

学校读者我要写书评

暂无评论

Numerical Simulation of Self-balanced Warm-Air Heating in Bo...

11th International Symposium on Heating, Ventilation and Air Conditioning, ISHVAC 2019

作者： Cao, Huizhe Wang, Yuhang Zhao, Tianli Qu, Jiagang Zhai, Jinfang Cai, Weihua School of Architecture Harbin Institute of Technology Harbin China Key Laboratory of Cold Region Urban and Rural Human Settlement Environment Science and Technology Ministry of Industry and Information Technology Harbin China School of Energy Science and Engineering Harbin Institute of Technology Harbin China Northeast Electric Power Design Institute China Power Engineering Consulting Group Changchun China School of Energy and Power Engineering Northeast Electric Power University Jilin China

ISBN: (纸本)9789811395192

The boiler-room of a large-scale thermal power plant belongs to super-high industrial building. The heat is concentrated in the upper part of the workshop. In the northern cold region, the wasted heat at the top of the boiler-room can be sent to the main plant-floor by the reversal of the roof-fans. The heat which dissipates from internal equipment is used to provide heating in winter, then the self-balanced heating can be achieved in the plant. In this paper, CFD simulation is used to analyze the high-pressure roof-fan reversing condition in the boiler-room. Combined with the airflow organization simulation results, different multiple airflow relay schemes are constructed through simulating the temperature field and velocity field to optimize the layout scheme of the relays of each equipment layer. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Roofs

Mid-infrared Chalcogenide Suspended Slot Waveguide for Gas Sensing 5

学校读者我要写书评

暂无评论

Mid-infrared Chalcogenide Suspended Slot Waveguide for Gas S...

5th Optoelectronics Global Conference, OGC 2020

作者： Pi, Mingquan Zheng, Chuantao Peng, Zihang Zhao, Huan Lang, Jiaming Liang, Lei Zhang, Yu Wang, Yiding Tittel, Frank K. Jilin University State Key Laboratory of Integrated Optoelectronics College of Electronic Science and Engineering 2699 Qianjin Street Changchun130012 China Changchun Institute of Optics Fine Mechanics and Physics Chinese Academy of Sciences State Key Laboratory of Luminescence and Applications Changchun130033 China Rice University Department of Electrical and Computer Engineering 6100 Main Street HoustonTX77005 United States

ISBN: (纸本)9781728198606

A chalcogenide (ChG) vertical slot waveguide and a ChG horizontal slot waveguide racetrack resonator were proposed for gas sensing. The vertical slot waveguide and the upper strip waveguide of the horizontal slot waveguide are suspended to enhance light-gas interaction. The waveguide sensors were optimized and the sensing performances were studied numerically. © 2020 IEEE.

关键词： Resonators

Transition between radial and toroidal orders in a trimer-based magnetic metasurface

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Tuz, Vladimir R. Evlyukhin, Andrey B. Fesenko, Volodymyr I. State Key Laboratory of Integrated Optoelectronics College of Electronic Science and Engineering International Center of Future Science Jilin University 2699 Qianjin St. Changchun130012 China School of Radiophysics Biomedical Electronics and Computer Systems V. N. Karazin Kharkiv National University 4 Svobody Sq. Kharkiv61022 Ukraine Institute of Quantum Optics Leibniz Universität Hannover Hannover30167 Germany Institute of Radio Astronomy of National Academy of Sciences of Ukraine 4 Mystetstv St. Kharkiv61002 Ukraine

The change in the arrangement of magnetic dipole moments in a magnetic metasurface, due to the influence of an external static magnetic field, is discussed. Each meta-atom of the metasurface is composed of three identical disk-shaped resonators (trimer) made of magnetically saturated ferrite. To provide physical insight, full-wave numerical simulations of the near-fields and transmission characteristics of the metasurface are complemented by the theoretical description based on the multipole decomposition method. With these methods, the study of eigenmodes and scattering conditions of a single magnetic resonator, trimer, and their array forming the metasurface is performed. It is found that the magnetic dipole-based collective hybrid mode of the trimer can be gradually transformed from the radial (pseudomonopole) to azimuthal (toroidal) order and vice versa by varying the bias magnetic field strength. This is because the magnetic dipole moment of each individual disk constituting the trimer undergoes rotation as the bias magnetic field strength changes. This transition between two orders is accompanied by various patterns of localization of the electric field inside the meta-atoms. Due to the unique field configuration of these modes, the proposed metasurface can be considered for designing magnetic field sensors and nonreciprocal devices. © 2023, CC BY.

关键词： Magnetic fields

Identifying Climatological Interactions in Global Influenza Across Temperate and Tropical Regions 9th

学校读者我要写书评

暂无评论

Identifying Climatological Interactions in Global Influenza ...

9th International Conference on Mobile Wireless Middleware, Operating Systems and Applications, MOBILWARE 2020

作者： Yuan, Zhilu Tang, Shengjun Huang, Qiuyang Zhang, Chijun Ertem, Zeynep Du, Zhanwei Bai, Yuan Research Institute for Smart Cities and Shenzhen Key Laboratory of Spatial Information Smart Sensing and Services School of Architecture and Urban Planning Shenzhen University Shenzhen China College of Computer Science and Technology Jilin University Changchun130012 China School of Management Science and Information Engineering Jilin University of Finance and Economics Changchun130117 China Marshall School of Business The University of Southerm California California United States Key Laboratory of Urban Land Resources Monitoring and Simulation MNR Shenzhen China

ISBN: (纸本)9783030622046

Recently, global epidemic models that uses climatological factors have been shown to explain influenza activities for both temperate and tropical regions. In this paper, we extend these global models by including interactions of climatological factors. We find that countries in Europe and Australia have higher forecast skill, indicating the stronger relationship of influenza with climatological factors, than regions in other continents. The influenza activities of 47 (83%) countries can be explained with a closer match using multi-factor interactions along with original factors than only using the original factors. © 2020, ICST institute for computer sciences, Social Informatics and Telecommunications engineering.

关键词： Tropics

Variant Plateau's Law in Atomically Thin Transition Metal Dichalcogenide Dome Networks

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Liu, Boqing Yildirim, Tanju Lü, Tieyu Blundo, Elena Wang, Li Jiang, Lixue Zou, Hongshuai Zhang, Lijun Zhao, Huijun Yin, Zongyou Tian, Fangbao Polimeni, Antonio Lu, Yuerui School of Engineering College of Engineering and Computer Science The Australian National University CanberraACT2601 Australia 1-1 Namiki Ibaraki Tsukuba305-0044 Japan Department of Physics Institute of Theoretical Physics and Astrophysics Xiamen University Xiamen361005 China Dipartimento di Fisica Sapienza Università di Roma Roma00185 Italy School of Engineering and Information Technology University of New South Wales CanberraACT2600 Australia Centre for Catalysis and Clean Energy Griffith University Gold Coast Campus QLD4222 Australia State Key Laboratory of Integrated Optoelectronics School of Materials Science and Engineering Jilin Provincial International Cooperation Key Laboratory of High-Efficiency Clean Energy Materials Jilin University Changchun130012 China Research School of Chemistry College of Science The Australian National University CanberraACT2601 Australia Arc Centre of Excellence in Quantum Computation and Communication Technology Anu Node CanberraACT2601 Australia

Since its fundamental inception from soap bubbles, Plateau's law has sparked extensive research in equilibrated states. However, most studies primarily relied on liquids, foams or cellular structures, whereas its applicability has yet to be explored in nano-scale solid films. Here, we observed a variant Plateau's law in networks of atomically thin domes made of solid two-dimensional (2D) transition metal dichalcogenides (TMDs). Discrete layer-dependent van der Waals (vdWs) interaction energies were experimentally and theoretically obtained for domes protruding in different TMD layers. Significant surface tension differences from layer-dependent vdWs interaction energies manifest in a variant of this fundamental law. Meanwhile, the remarkable mechanical properties, gas impermeability and interlayer vdWs interaction energy of TMD films enable domes and the networks to sustain high gas pressure and exist in a fundamentally variant nature for several years. Our findings pave the way towards exploring variant discretised states with applications in opto-electro-mechanical devices. Copyright © 2023, The Authors. All rights reserved.

关键词： Surface tension

Opening of Band Gap of Graphene with High Electronic Mobility by Codoping BN Pairs

学校读者我要写书评

暂无评论

Chemical Research in Chinese Universities 2019年第6期35卷 1058-1061页

作者： REN Xiangyang XIA Sha ZHANG Zhiguo MENG Xing YU Hongmei WU Qi ZHANG Wenyi LI Aiwu YANG Han State Key Laboratory of Integrated Optoelectronics College of Electronic Science and EngineeringJilin UniversityChangchun 130012P.R.China Shenyang Acadamy of Instrumentation Science Co. Ltd.Shenyang 110043P.R.China College of Physics Jilin UniversityChangchun 130012P.R.China College of Computer Science and Technology Jilin UniversityChangchun 130012P.R.China Systems Engineering Research Institute Beijing 100036P.R.China

Two-dimensional(2D)materials with a high density and low power consumption have become the most popular candidates for next-generation semiconductor electronic *** a prototype 2D material,graphene has attracted much attention owing to its stability and ultrahigh ***,zero band gap of graphene leads to very low on-off ratios and thus limits its applications in electronic devices,such as *** some new 2D materials and doped graphene have nonzero band gaps,the electronic mobility is *** this study,to open the band gap of graphene with high electronic mobility,the structure and property of BN-doped graphene were evaluated using first-principles *** formation energies indicate that the six-membered BN rings doped graphene has the most favorable *** band structures show that the band gaps can be opened by such type of ***,the Dirac-cone-like band dispersion of graphene is mostly inhibited,ensuring high electronic ***,codoping BN into graphene might provide 2D materials with nonzero band gaps and high electronic mobility.

关键词： First-principle calculation Doping Optoeletronic property Mobility

A block-based generative model for attributed networks embedding

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Liu, Xueyan Song, Wenzhuo Zuo, Wanli Musial, Katarzyna Yang, Bo School of Computer Science and Technology Jilin University Changchun Jilin130012 China Ministry of Education Changchun Jilin130012 China Advanced Analytics Institute School of Computer Science Faculty of Engineering and IT University of Technology Sydney NSW2007 Australia

Attributed network embedding has attracted plenty of interests in recent years. It aims to learn task-independent, low-dimension, and continuous vectors for nodes preserving both topology and attribute information. Most existing methods, such as GCN and its variations, mainly focus on the local information, i.e., the attributes of the neighbors. Thus, they have been well studied for assortative networks but ignored disassortative networks, which are common in real scenes. To address this issue, we propose a block-based generative model for attributed network embedding on a probability perspective inspired by the stochastic block model (SBM). Specifically, the nodes are assigned to several blocks wherein the nodes in the same block share the similar link patterns. These patterns can define assortative networks containing communities or disassortative networks with the multipartite, hub, or any hybrid structures. Concerning the attribute information, we assume that each node has a hidden embedding related to its assigned block, and then we use a neural network to characterize the nonlinearity between the node embedding and its attribute. We perform extensive experiments on real-world and synthetic attributed networks, and the experimental results show that our proposed method remarkably outperforms state-of-the-art embedding methods for both clustering and classification tasks, especially on disassortative networks. Copyright © 2020, The Authors. All rights reserved.

关键词： Network embeddings

Covert Prompt Transmission for Secure Large Language Model Services

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zhang, Ruichen Liu, Yinqiu Tang, Shunpu Wang, Jiacheng Niyato, Dusit Sun, Geng Li, Yonghui Sun, Sumei College of Computing and Data Science Nanyang Technological University Singapore College of Computer Science and Technology Jilin University Changchun130012 China School of Electrical and Information Engineering University of Sydney SydneyNSW2006 Australia Institute for Infocomm Research Agency for Science Technology and Research Singapore

This paper investigates covert prompt transmission for secure and efficient large language model (LLM) services over wireless networks. We formulate a latency minimization problem under fidelity and detectability constraints to ensure confidential and covert communication by jointly optimizing the transmit power and prompt compression ratio. To solve this problem, we first propose a prompt compression and encryption (PCAE) framework, performing surprisal-guided compression followed by lightweight permutation-based encryption. Specifically, PCAE employs a locally deployed small language model (SLM) to estimate token-level surprisal scores, selectively retaining semantically critical tokens while discarding redundant ones. This significantly reduces computational overhead and transmission duration. To further enhance covert wireless transmission, we then develop a group-based proximal policy optimization (GPPO) method that samples multiple candidate actions for each state, selecting the optimal one within each group and incorporating a Kullback-Leibler (KL) divergence penalty to improve policy stability and exploration. Simulation results show that PCAE achieves comparable LLM response fidelity to baseline methods while reducing preprocessing latency by over five orders of magnitude, enabling real-time edge deployment. We further validate PCAE effectiveness across diverse LLM backbones, including DeepSeek-32B, Qwen-32B, and their smaller variants. Moreover, GPPO reduces covert transmission latency by up to 38.6% compared to existing reinforcement learning strategies, with further analysis showing that increased transmit power provides additional latency benefits. © 2025, CC BY.

关键词： Deep reinforcement learning