检索结果-内蒙古大学图书馆

Lay CO:Achieving Least Lossy Accuracy for Most Efficient RRAM-Based Deep Neural Network Accelerator via Layer-Centric Co-Optimization

引用

计算机科学技术学报（英文版） 2023年第2期38卷 328-347页

作者： Shao-Feng Zhao Fang Wang Bo Liu Dan Feng Yang Liu Wuhan National Laboratory for Optoelectronics Huazhong University of Science and TechnologyWuhan 430074China School of Computer Science and Technology Huazhong University of Science and TechnologyWuhan 430074China Cloud Computing and Big Data Institute Henan University of Economics and LawZhengzhou 450001China Wuhan National Laboratory for Optoelectronics Huazhong University of Science and TechnologyWuhan 430074China School of Computer Science and Technology Huazhong University of Science and TechnologyWuhan 430074China Research Institute of Huazhong University of Science and Technology in Shenzhen Shenzhen 518057China School of Computer and Artificial Intelligence Zhengzhou UniversityZhengzhou 450001China Wuhan National Laboratory for Optoelectronics Huazhong University of Science and TechnologyWuhan 430074China School of Computer Science and Technology Huazhong University of Science and TechnologyWuhan 430074China Cloud Computing and Big Data Institute Henan University of Economics and LawZhengzhou 450001China

Resistive random access memory(RRAM)enables the functionality of operating massively parallel dot prod-ucts and ***-based accelerator is such an effective approach to bridging the gap between Internet of Things devices'constrained resources and deep neural networks'tremendous *** to the huge overhead of Analog to Digital(A/D)and digital accumulations,analog RRAM buffer is introduced to extend the processing in analog and in *** analog RRAM buffer offers potential solutions to A/D conversion issues,the energy consumption is still challenging in resource-constrained environments,especially with enormous intermediate data ***,criti-cal concerns over endurance must also be resolved before the RRAM buffer could be frequently used in reality for DNN in-ference *** we propose LayCO,a layer-centric co-optimizing scheme to address the energy and endurance con-cerns altogether while strictly providing an inference accuracy *** relies on two key ideas:1)co-optimizing with reduced supply voltage and reduced bit-width of accelerator architectures to increase the DNN's error tolerance and achieve the accelerator's energy efficiency,and 2)efficiently mapping and swapping individual DNN data to a correspond-ing RRAM partition in a way that meets the endurance *** evaluation with representative DNN models demonstrates that LayCO outperforms the baseline RRAM buffer based accelerator by 27x improvement in energy effi-ciency(over TIMELY-like configuration),308x in lifetime prolongation and 6x in area reduction(over RAQ)while main-taining the DNN accuracy loss less than 1％.

关键词： deep learning accelerator resistive random access memory (RRAM) energy efficiency error tolerance

来源：评论

学校读者我要写书评

暂无评论

Vision Intelligence Assisted Lung Function Estimation Based on Transformer Encoder-Decoder Network With Invertible Modeling

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第7期5卷 3336-3349页

作者： Chen, Liuyin Lu, Di Zhai, Jianxue Cai, Kaican Wang, Long Zhang, Zijun City University of HongKong School of Data Science Hong Kong Nanfang Hospital Southern Medical University Department of Thoracic Surgery Guangzhou510515 China University of Science and Technology Beijing Department of Computer Science and Technology Beijing100083 China University of Science and Technology Beijing Shunde Graduate School Foshan China

Lung function evaluation is important to many medical applications, but conducting pulmonary function tests is constrained by different conditions. This article presents a pioneer study of an integrated invertible deep learning method for lung function estimation via using computed tomography (CT) images. First, the projection method is proposed to flatten the three-dimensional (3-D) image onto a two-dimensional (2-D) plane, with preserving location information in 3-D. Next, the MBConv transformer-based encoder-decoder structure is developed to extract latent features. Finally, we develop an invertible normalizing flow (NF) model to infer lung function based on the extracted features and design two loss functions for two directions. The method enables both estimating the lung function based on CT images and metadata as well as generating the corresponding simulated CT image according to the lung function. Computational studies show that the proposed regression model outperforms all state-of-the-art image regression models. A comprehensive comparative analysis also demonstrates the effectiveness of using generated images and confirms the superiority of the proposed method. To the best of our knowledge, this work is the first of its kind in combining encoder-decoder network with NFs to ensure the effectiveness of the fully invertible framework, especially in lung CT image analysis. © 2020 IEEE.

关键词： Metadata

来源：评论

学校读者我要写书评

暂无评论

Latency- and Privacy-Aware Convolutional Neural Network Distributed Inference for Reliable Artificial Intelligence Systems

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第2期6卷 365-377页

作者： Hu, Yuhao Xu, Xiaolong Qi, Lianyong Zhou, Xiaokang Xia, Xiaoyu Nanjing University of Information Science and Technology School of Software Nanjing210044 China College of computer science and technology Qingdao266850 China Shiga University Faculty of Data Science Hikone522-8522 Japan RMIT University School of Computing Technologies MelbourneVIC3000 Australia

Reliable artificial intelligence (AI) systems not only propose a challenge on providing intelligent services with high quality for customers but also require customers' privacy to be protected as much as possible during process of the services. Given the ultrahigh computing load brought by deep-learning-based intelligent services to edge devices and the ultralong distance between edge and cloud, low-latency requirement of intelligent services is hard to meet in single edge computing or cloud computing. Edge-cloud collaborative inference of deep neural networks (DNNs) is considered a feasible solution to the problem. However, former work has not reduced the inference latency to the greatest extent and has not considered privacy protection in distributed systems. To solve the problem, we first establish a novel queue mechanism. Then, the convolution layer split decisions are made based on deep reinforcement learning (DRL) to realize the parallel inference of convolutional neural networks (CNNs) for inference latency reduction. Next, for each CNN, the partition decision is made based on brute force algorithm to further reduce inference latency and protect customers' privacy. Finally, simulation results show that our method performs better than the existing other methods. © 2020 IEEE.

关键词： Cloud computing

来源：评论

学校读者我要写书评

暂无评论

DP-PartFIM: Frequent Itemset Mining using Differential Privacy and Partition

引用

IEEE Transactions on Emerging Topics in Computing 2024年 1-11页

作者： Liu, Xinyu Gan, Wensheng Yu, Lele Liu, Yining School of Data Science and Artificial Intelligence Wenzhou University of Technology Wenzhou China College of Cyber Security Jinan University Guangzhou China

Itemset mining is a popular data mining technique for extracting interesting and valuable information from large datasets. However, since datasets contain sensitive private data, it is not permitted to directly mine the data or share the mining results. Previous privacy-preserving frequent itemset mining research was not efficient because of the use of privacy budgets or long transaction truncation strategies, which are impractical for large datasets. In this paper, we propose a more efficient partition mining technology, DP-PartFIM, based on differential privacy, which protects privacy while mining data. DP-PartFIM uses partition mining to mine frequent itemsets and constructs vertical data storage formats for each partition, which makes the algorithm equally efficient for large datasets. To protect data privacy, DP-PartFIM adds Laplace noise to support candidate itemsets. The experimental results show that, compared with the classical privacy-preserving itemset mining methods, DP-PartFIM better guarantees data utility and privacy. IEEE

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

ERROR ANALYSIS OF VIRTUAL ELEMENT METHODS FOR THE TIME-DEPENDENT POISSON-NERNST-PLANCK EQUATIONS

引用

Journal of Computational Mathematics 2025年第3期43卷 731-770页

作者： Ying Yang Ya Liu Yang Liu Shi Shu School of Mathematics and Computational Science Guangxi Colleges and Universities Key Laboratory of Data Analysis and ComputationCenter for Applied Mathematics of Guangxi(GUET)Guilin University of Electronic TechnologyGuilin 541004China School of Mathematics and Computational Science Guilin University of Electronic TechnologyGuilin 541004China School of Mathematics and Computational Science Xiangtan UniversityXiangtan 411105China

We discuss and analyze the virtual element method on general polygonal meshes for the time-dependent Poisson-Nernst-Planck(PNP)equations,which are a nonlinear coupled system widely used in semiconductors and ion *** presenting the semi-discrete scheme,the optimal H1 norm error estimates are presented for the time-dependent PNP equations,which are based on some error estimates of a virtual element energy *** Gummel iteration is used to decouple and linearize the PNP equations and the error analysis is also given for the iteration of fully discrete virtual element *** numerical experiment on different polygonal meshes verifies the theoretical convergence results and shows the efficiency of the virtual element method.

关键词： Virtual element method Error estimate Poisson-Nernst-Planck equations Polygonal meshes Energy projection Gummel iteration

来源：评论

学校读者我要写书评

暂无评论

Key-Value Store Coupled with an Operating System for Storing Large-Scale Values

引用

Computers, Materials & Continua 2022年第11期73卷 3333-3350页

作者： Jeonghwan Im Hyuk-Yoon Kwon Graduate School of Data Science Seoul National University of Science and TechnologySeoulKorea Department of Industrial Engineering Graduate School of Data ScienceResearch Center for Electrical and Information ScienceSeoul National University of Science and TechnologySeoulKorea

The key-value store can provide flexibility of data types because it does not need to specify the data types to be stored in advance and can store any types of data as the value of the key-value *** types of studies have been conducted to improve the performance of the key-value store while maintaining its ***,the research efforts storing the large-scale values such as multimedia data files(e.g.,images or videos)in the key-value store were *** this study,we propose a new key-value store,WR-Store++aiming to store the large-scale values ***,it provides a new design of separating data and index by working with the built-in data structure of the Windows operating system and the file *** utilization of the built-in data structure of the Windows operating system achieves the efficiency of the key-value store and that of the file system extends the limited space of the storage *** also present chunk-based memory management and parallel processing of WR-Store++to further improve its performance in the GET *** the experiments,we show that WR-Store++can store at least 32.74 times larger datasets than the existing baseline key-value store,WR-Store,which has the limitation in storing large-scale data ***,in terms of processing efficiency,we show that WR-Store++outperforms not only WR-Store but also the other state-ofthe-art key-value stores,LevelDB,RocksDB,and BerkeleyDB,for individual key-value operations and mixed workloads.

关键词： Key-value stores large-scale values chunk-based memory management parallel processing

来源：评论

学校读者我要写书评

暂无评论

A blockchain-empowered authentication scheme for worm detection in wireless sensor network

引用

Digital Communications and Networks 2024年第2期10卷 265-272页

作者： Yuling Chen Xiong Yang Tao Li Yi Ren Yangyang Long State Key Laboratory of Public Big Data College of Computer Science and TechnologyGuizhou UniversityGuiyang550025China School of Computing Science University of East AngliaNorwichNR57TJUK

Wireless Sensor Network(WSN)is a distributed sensor network composed a large number of nodes with low cost,low performance and *** special structure of WSN brings both convenience and *** example,a malicious participant can launch attacks by capturing a physical ***,node authentication that can resist malicious attacks is very important to network ***,blockchain technology has shown the potential to enhance the security of the Internet of Things(IoT).In this paper,we propose a Blockchain-empowered Authentication Scheme(BAS)for *** our scheme,all nodes are managed by utilizing the identity information stored on the ***,the simulation experiment about worm detection is executed on BAS,and the security is evaluated from detection and infection *** experiment results indicate that the proposed scheme can effectively inhibit the spread and infection of worms in the network.

关键词： Wireless Sensor Network(WSN) Node authentication Blockchain Tangle Worm detection

来源：评论

学校读者我要写书评

暂无评论

An electron-transfer-tuning strategy at the graphene/metal interface for improving acidic water electrolysis in proton exchange membrane electrolyzers

引用

Journal of Energy Chemistry 2025年第4期103卷 344-352页

作者： Yue Xu Shuaidong Li Yingjian He Fumiya Shiokawa Samuel Jeong Aimi Asilah Haji Tajuddin Zeyu Xi Yoshikazu Ito Jingzi Zhang Zeyun Cai Xi Lin Kailong Hu School of Materials Science and Engineering and Institute of Materials Genome&Big DataHarbin Institute of TechnologyShenzhen 518055GuangdongChina Institute of Special Environments Physical Sciences Harbin Institute of TechnologyShenzhen 518055GuangdongChina School of Computer Science and Technology Harbin Institute of TechnologyShenzhen 518055GuangdongChina Institute of Applied Physics Graduate School of Pure and Applied SciencesUniversity of TsukubaTsukuba 305-8573Japan

Graphene encapsulation has been shown to be an effective technique for improving the corrosion resistance of non-noble metal catalysts for the acidic water *** key challenge lies in enhancing the electrocatalytic activity of graphene-encapsulated metals while maintaining their durability in acidic ***,an electron-transfer-tuning strategy is investigated at the graphene/NiMo interface,aiming to improve the hydrogen evolution reaction(HER)performance of graphene-encapsulated NiMo *** doping of Ti,a low electronegativity element,into NiMo substrate was confirmed to increase electron transfer from the metal core toward the *** electron-rich state on graphene facilitates the adsorption of positively charged protons on graphene,thereby enabling a Pt/C-comparable performance in 0.5 M H_(2)SO_(4),with only a 3.8%degradation in performance over a 120-h continuous *** proton exchange membrane(PEM)water electrolyzer assembled by the N-doped grapheneencapsulated Ti-doped NiMo exhibits a smaller cell voltage to achieve a current density of 2.0 A cm^(-2),in comparison to the Pt/C based *** study proposes a novel electron-transfer-tuning strategy to improve the HER activity of graphene-encapsulated non-noble metal catalysts without sacrificing durability in acidic electrolytes.

关键词： Hydrogen evolution Acidic media Graphene encapsulation Interface Proton exchange membrane

来源：评论

学校读者我要写书评

暂无评论

Diluie: constructing diverse demonstrations of in-context learning with large language model for unified information extraction

引用

Neural Computing and Applications 2024年第22期36卷 13491-13512页

作者： Guo, Qian Guo, Yi Zhao, Jin Department of Computer Science and Engineering East China University of Science and Technology Shanghai200237 China Shanghai Key Laboratory of Data Science School of Computer Science Fudan University Shanghai200433 China

Large language models (LLMs) have demonstrated promising in-context learning capabilities, especially with instructive prompts. However, recent studies have shown that existing large models still face challenges in specific information extraction (IE) tasks. Moreover, it could have more effectively utilized various prompts such as instruction tuning, diverse demonstrations of in-context learning, and long-range token sequences for assisting language modeling in understanding context. In this study, we propose DILUIE, a unified information extraction framework based on in-context learning with diverse demonstration examples. DILUIE is encoded with an EVA attention mechanism and incremental encoding technology. Based on the constructed diverse demonstrations, we expand the size of instances efficiently in both instruction tuning and in-context learning to gain insights into the potential benefits of utilizing diverse information extraction datasets. To deepen the understanding of context, we further design three auxiliary tasks to assist in aligning contextual semantics. Experimental results demonstrate that DILUIE achieves 2.23 and 2.53% improvements in terms of Micor-/Macor-F1 on average relative to the current state-of-the-art baseline, which also significantly outperforms the GPT-3.5-turbo in zero-shot settings, and the average token length of achieving the best performance over tasks is around 15k. Furthermore, we observe that in-context learning shows enhanced performance when provided with more demonstrations during multiple-shot instruction tuning (8 k). Additionally, increasing the length of instructions (10 k) can result in a more substantial improvement in the upper limits of scaling for in-context learning. Code is available on https://***/Phevos75/DILUIE. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Demonstrations

来源：评论

学校读者我要写书评

暂无评论

Passively Q-switched operation of a 1.94μm thulium-doped solid-state laser based on MXeneV_(2)CT_(x)

引用

Chinese Optics Letters 2023年第2期21卷 40-44页

作者：袁晋鹤李佳睿吴政阳李双成韩杰李林军 College of Physical Science and Technology Heilongjiang UniversityHarbin 150080China Heilongjiang Provincial Key Laboratory of Laser Spectroscopy Technology and Application Harbin University of Science and TechnologyHarbin 150080China School of Data Science and Technology Heilongjiang UniversityHarbin 150080China

MXene V_(2)CT_(x) has great practicability because it is not easy to degrade under ambient *** this paper,a V_(2)CT_(x) saturable absorber(SA)was firstly applied to a passively Q-switched(PQS)laser,to the best of our *** V_(2)CT_(x)-SA was prepared by the spin-coating *** linear absorption of the V_(2)CT_(x)-SA in the 1000-2200 nm region and the nonlinear absorption near 2μm were *** the V_(2)CT_(x)-SA,a typical PQS operation at 1.94μm was realized in a Tm:*** minimum pulse width produced by the PQS laser was 528 ns,and the peak power,repetition rate,and average output power were 10.06 W,65.9 kHz,and 350 mW,***,the maximum pulse energy was 6.33μ*** work demonstrates that the V_(2)CT_(x) can be used as an effective SA to obtain nanosecond pulses with high peak power and high repetition rate simultaneously.

关键词： passive Q-switching MXene V2CTx saturable absorber thulium doping

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：