检索结果-内蒙古大学图书馆

19th International Conference on Advanced Data Mining and Applications, ADMA 2023

作者： Chen, Zhihao Wang, Hongya School of Computer Science and Technology Donghua University Shanghai China State Key Laboratory of Computer Architecture Institute of Computing Technology CAS Beijing China Shanghai Key Laboratory of Computer Software Evaluating and Testing Shanghai China

ISBN: (纸本)9783031466731

Currently, large-scale vision and language models has significantly improved the performances of cross-modal retrieval tasks. However, large-scale models require a substantial amount of computing resources, so the execution of these models on devices with limited resources is challenging. Thus, it is paramount to reduce the model size and minimize computing costs of a model without sacrificing its performance. In this paper, we improved TERAN by dividing cross-modal retrieval into two stages: image-text coarse-grained matching and image-text fine-grained matching. Specifically, we present a novel approach called Two-Stage Cross-Modal Retrieval network(TSCMR). To reduce model size after model training, our approach utilized a new knowledge distillation method for Transformer-based models. Experiments have shown that our approach maintains a performance comparable to TERAN on the MS-COCO 1K test set, while being 2x smaller and 3.1x faster on inference. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

UKF‐MOT:An unscented Kalman filter‐based 3D multi‐object tracker

引用

CAAI Transactions on Intelligence technology 2024年第4期9卷 1031-1041页

作者： Meng Liu Jianwei Niu Yu Liu Collective Intelligence&Collaboration Laboratory China North Artificial Intelligence and Innovation Research InstituteBeijingChina China North Vehicle Research Institute BeijingChina State Key Laboratory of Software Development Environment School of Computer Science and EngineeringBeihang UniversityBeijingChina State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBeihang UniversityBeijingChina School of Computer Science and Engineering Beihang UniversityBeijingChina

Multi‐object tracking in autonomous driving is a non‐linear *** better address the tracking problem,this paper leveraged an unscented Kalman filter to predict the object's *** the association stage,the Mahalanobis distance was employed as an affinity metric,and a Non‐minimum Suppression method was designed for *** the detections fed into the tracker and continuous‘predicting‐matching’steps,the states of each object at different time steps were described as their own continuous *** conducted extensive experiments to evaluate tracking accuracy on three challenging datasets(KITTI,nuScenes and Waymo).The experimental results demon-strated that our method effectively achieved multi‐object tracking with satisfactory ac-curacy and real‐time efficiency.

关键词： autonomous vehicle transportation

来源：评论

学校读者我要写书评

暂无评论

AI Computing Systems for Large Language Models Training

引用

计算机科学技术学报（英文版） 2025年第1期40卷 6-41页

作者：张振兴文渊博吕涵祺刘畅张蕊李夏青王超杜子东郭崎李玲周学海陈云霁 1School of Computer Science and Technology University of Science and Technology of China Hefei 230026 China 2State Key Laboratory of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing 100190 China 3Cambricon Technologies Beijing 100191 China 4Shanghai Innovation Center for Processor Technologies Shanghai 201210 China 5Intelligent Software Research Center Institute of Software Chinese Academy of Sciences Beijing 100190 China 6University of Chinese Academy of Sciences Beijing 101408 China

In this paper,we present a comprehensive overview of artificial intelligence(AI)computing systems for large language models(LLMs)*** rapid advancement of LLMs in recent years,coupled with the widespread adoption of algorithms and applications such as BERT,ChatGPT,and DeepSeek,has sparked significant interest in this *** classify LLMs into encoder-only,encoder-decoder,and decoder-only models,and briefly analyze their training and infer-ence processes to emphasize their substantial need for computational *** operations depend heavily on AI-specific accelerators like GPUs(graphics processing units),TPUs(tensor processing units),and MLUs(machine learning units).However,as the gap widens between the increasing complexity of LLMs and the current capabilities of accelerators,it becomes essential to adopt heterogeneous computing systems optimized for distributed environments to manage the growing computational and memory requirements of *** delve into the execution and scheduling of LLM algo-rithms,underlining the critical role of distributed computing strategies,memory management enhancements,and boosting computational *** paper clarifies the complex relationship between algorithm design,hardware infrastructure,and software optimization,and provides an in-depth understanding of both the software and hardware infrastructure sup-porting LLMs training,offering insights into the challenges and potential avenues for future development and deployment.

关键词： artificial intelligence(AI)chip large language model(LLM) AI computing system accelerator

来源：评论

学校读者我要写书评

暂无评论

piRT-IFC:Physics-informed real-time impedance flow cytometry for the characterization of cellular intrinsic electrical properties

引用

Microsystems & Nanoengineering 2023年第3期9卷 347-356页

作者： Xiaofeng Luan Pengbin Liu Di Huang Haiping Zhao Yuang Li Sheng Sun Wenchang Zhang Lingqian Zhang Mingxiao Li Tian Zhi Yang Zhao Chengjun Huang Institute of Microelectronics of the Chinese Academy of Sciences BeijingChina University of Chinese Academy of Sciences BeijingChina State Key Laboratory of Computer Architecture Institute of Computing TechnologyChinese Academy of SciencesBeijingChina Cerebrovascular Diseases Research Institute Xuanwu Hospital of Capital Medical UniversityBeijingChina

Real-time transformation was important for the practical implementation of impedance flow *** major obstacle was the time-consuming step of translating raw data to cellular intrinsic electrical properties(e.g.,specific membrane capacitance C_(sm) and cytoplasm conductivityσ_(cyto)).Although optimization strategies such as neural network-aided strategies were recently reported to provide an impressive boost to the translation process,simultaneously achieving high speed,accuracy,and generalization capability is still *** this end,we proposed a fast parallel physical fitting solver that could characterize single cells’C_(sm)andσ_(cyto)within 0.62 ms/cell without any data preacquisition or pretraining *** achieved the 27000-fold acceleration without loss of accuracy compared with the traditional *** on the solver,we implemented physics-informed real-time impedance flow cytometry(piRT-IFC),which was able to characterize up to 100,902 cells’C_(sm) andσ_(cyto)within 50 min in a real-time *** to the fully connected neural network(FCNN)predictor,the proposed real-time solver showed comparable processing speed but higher ***,we used a neutrophil degranulation cell model to represent tasks to test unfamiliar samples without data for *** being treated with cytochalasin B and N-Formyl-Met-Leu-Phe,HL-60 cells underwent dynamic degranulation processes,and we characterized cell’s C_(sm)andσ_(cyto)using *** to the results from our solver,accuracy loss was observed in the results predicted by the FCNN,revealing the advantages of high speed,accuracy,and generalizability of the proposed piRT-IFC.

关键词： impedance intrinsic flow

来源：评论

学校读者我要写书评

暂无评论

Survey on digital twins for Internet of Vehicles:Fundamentals,challenges,and opportunities

引用

Digital Communications and Networks 2024年第2期10卷 237-247页

作者： Jiajie Guo Muhammad Bilal Yuying Qiu Cheng Qian Xiaolong Xu Kim-Kwang Raymond Choo School of Computer and Software Nanjing University of Information Science and TechnologyNanjing210044China Department of Computer Engineering Hankuk University of Foreign StudiesGyeonggi-do17035South Korea Jiangsu Hydraulic Research Institute Nanjing210044China Department of Information Systems and Cyber Security University of Texas at San AntonioSan AntonioTX78249-0631USA

As autonomous vehicles and the other supporting infrastructures(e.g.,smart cities and intelligent transportation systems)become more commonplace,the Internet of Vehicles(IoV)is getting increasingly *** have been attempts to utilize Digital Twins(DTs)to facilitate the design,evaluation,and deployment of IoV-based systems,for example by supporting high-fidelity modeling,real-time monitoring,and advanced predictive ***,the literature review undertaken in this paper suggests that integrating DTs into IoV-based system design and deployment remains an understudied *** addition,this paper explains how DTs can benefit IoV system designers and implementers,as well as describes several challenges and opportunities for future researchers.

关键词： Internet of vehicles Digital twin Simulation Traffic systems

来源：评论

学校读者我要写书评

暂无评论

A Tutorial on Federated Learning from Theory to Practice:Foundations,software Frameworks,Exemplary Use Cases,and Selected Trends

引用

IEEE/CAA Journal of Automatica Sinica 2024年第4期11卷 824-850页

作者： M.Victoria Luzón Nuria Rodríguez-Barroso Alberto Argente-Garrido Daniel Jiménez-López Jose M.Moyano Javier Del Ser Weiping Ding Francisco Herrera Department of Software Engineering Andalusian Research Institute in Data Science and Computational Intelligence(DaSCI)University of GranadaGranada 18071Spain Department of Computer Science and Artificial Intelligence Andalusian Research Institute in Data Science and Computational Intelligence(DaSCI)University of GranadaGranada 18071Spain Department of Communications Engineering University of the Basque Country(UPV/EHU)and also with TECNALIABasque Research&Technology Alliance(BRTA)Spain School of Information Science and Technology Nantong UniversityNantong 226019China

When data privacy is imposed as a necessity,Federated learning(FL)emerges as a relevant artificial intelligence field for developing machine learning(ML)models in a distributed and decentralized *** allows ML models to be trained on local devices without any need for centralized data transfer,thereby reducing both the exposure of sensitive data and the possibility of data interception by malicious third *** paradigm has gained momentum in the last few years,spurred by the plethora of real-world applications that have leveraged its ability to improve the efficiency of distributed learning and to accommodate numerous participants with their data *** virtue of FL,models can be learned from all such distributed data sources while preserving data *** aim of this paper is to provide a practical tutorial on FL,including a short methodology and a systematic analysis of existing software ***,our tutorial provides exemplary cases of study from three complementary perspectives:i)Foundations of FL,describing the main components of FL,from key elements to FL categories;ii)Implementation guidelines and exemplary cases of study,by systematically examining the functionalities provided by existing software frameworks for FL deployment,devising a methodology to design a FL scenario,and providing exemplary cases of study with source code for different ML approaches;and iii)Trends,shortly reviewing a non-exhaustive list of research directions that are under active investigation in the current FL *** ultimate purpose of this work is to establish itself as a referential work for researchers,developers,and data scientists willing to explore the capabilities of FL in practical applications.

关键词： Data privacy distributed machine learning federated learning software frameworks

来源：评论

学校读者我要写书评

暂无评论

Accelerating Hierarchical Federated Learning with Model Splitting in Edge Computing 30

Accelerating Hierarchical Federated Learning with Model Spli...

引用

30th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2024

作者： Wang, Xiangnan Xu, Yang Xu, Hongli Sun, Zhipeng Liao, Yunming Qi, Ji University of Science and Technology of China School of Computer Science and Technology Hefei China University of Science and Technology of China Suzhou Institute for Advanced Research Suzhou China Software Technology Co. Ltd China

ISBN: (纸本)9798331515966

Recently, Hierarchical Federated Learning (HFL) stands out as a cutting-edge approach to efficiently learn knowledge from massive data on edge devices or clients. To alleviate the computation/communication burden of training large-scale models on resource-constrained clients, Hierarchical Split Federated Learning (HSFL), which splits an entire model into a top and bottom (sub-)model and offloads the training of the top model to the edge server, has been proposed. Nonetheless, there are two key issues, i.e., system heterogeneity and dynamic contexts, hindering the application of effective HSFL. In response, we present an efficient HSFL method, termed AdaHSFL, which introduces intra-cluster gradient feedback regulation and inter-cluster updating frequencies optimization to enhance training efficiency. Concretely, intra-cluster gradient feedback regulation enables edge servers to instantly process incoming smashed data and calculate gradients using top model copies on background threads, while inter-cluster updating frequencies optimization adjusts edge cluster updating frequencies to align the training duration across all clusters with that of the fastest cluster. Furthermore, AdaHSFL explores to jointly implement these two strategies to eliminate the idle waiting time both intra- and intercluster incurred by synchronization barriers. Rigorous performance evaluation demonstrates that AdaHSFL can improve the accuracy by 3.9%-25.5% within given time budgets, compared with the baselines. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Global video object segmentation with spatial constraint module

引用

Computational Visual Media 2023年第2期9卷 385-400页

作者： Yadang Chen Duolin Wang Zhiguo Chen Zhi-Xin Yang Enhua Wu Engineering Research Center of Digital Forensics Ministry of EducationSchool of Computer and SoftwareNanjing University of Information Science and TechnologyNanjing 210044China State Key Laboratory of Internet of Things for Smart City Department of Electromechanical EngineeringUniversity of MacaoMacao 999078China State Key Laboratory of Computer Science Institute of SoftwareUniversity of Chinese Academy of SciencesBeijing 100190China Faculty of Science and Technology University of MacaoMacao 999078China

We present a lightweight and efficient semisupervised video object segmentation network based on the space-time memory *** some extent,our method solves the two difficulties encountered in traditional video object segmentation:one is that the single frame calculation time is too long,and the other is that the current frame’s segmentation should use more information from past *** algorithm uses a global context(GC)module to achieve highperformance,real-time *** GC module can effectively integrate multi-frame image information without increased memory and can process each frame in real ***,the prediction mask of the previous frame is helpful for the segmentation of the current frame,so we input it into a spatial constraint module(SCM),which constrains the areas of segments in the current *** SCM effectively alleviates mismatching of similar targets yet consumes few additional *** added a refinement module to the decoder to improve boundary *** model achieves state-of-the-art results on various datasets,scoring 80.1%on YouTube-VOS 2018 and a J&F score of 78.0%on DAVIS 2017,while taking 0.05 s per frame on the DAVIS 2016 validation dataset.

关键词： video object segmentation semantic segmentation global context(GC)module spatial constraint

来源：评论

学校读者我要写书评

暂无评论

Cost/Benefit Analysis of AIoT Image Sensing for Construction Safety Monitoring

引用

Journal of Engineering, Project, and Production Management 2024年第3期14卷 1-13页

作者： Wang, Rong-Jing Yu, Wen-Der Liao, Hsien-Chou Chang, Hsien-Kuan Lim, Zi-Yi Architecture and Building Research Institute Ministry of Interior New Taipei23143 Taiwan Department of Construction Engineering Chaoyang University of Technology Taichung413310 Taiwan Department of Computer Science and Information Technology Chaoyang University of Technology Taichung413310 Taiwan Department of Information and Communication Engineering Chaoyang University of Technology Taichung413310 Taiwan

Rapid advances in deep learning and computer vision enable traditional cloud-based decision-making through edge computing with the Artificial Intelligent Internet of Things (AIoT) image sensors (AIoT-IS), thus improving the timeliness and security of image recognition. This study is indented to investigate the potential costs and benefits of AIoT-IS applications. This study summarizes AIoT-IS application scenarios for construction safety monitoring and proposes a cost/benefit analysis method for AIoT-IS implementation projects. According to the case study results, AIoT-IS achieves significant benefits, with a Net Present Value Index (NPVI) of 19.17% and a Benefit/Cost Ratio (BCR) of 4.65 as applied to construction site safety monitoring. Interviews with domain experts also provided qualitative feedback, pointing to the directions for future research. The proposed method is applicable for the decision-making of AIoT-IS adoption and the feasibility assessment of other innovative construction technologies. Copyright © Journal of Engineering, Project, and Production Management (EPPM-Journal).

关键词： Cost benefit analysis

来源：评论

学校读者我要写书评

暂无评论

research on Imputation Method for Missing Values of Nuclear Power Plants' Operation Data 13

Research on Imputation Method for Missing Values of Nuclear ...

引用

13th International Conference on Electronics, Communications and Networks, CECNet 2023

作者： Liu, Jie Qi, Xiaodi Wu, Zhiqiang Jiang, Wei Chen, Zhi Jian, Yifan Wen, Juan School of Computer/Software University of South China China Laboratory of Nuclear Reactor System Design Technology China Nuclear Power Research and Design Institute China

ISBN: (纸本)9781643684802

The collection of nuclear power plants operating data is the basis for subsequent fault diagnosis and obtaining the operating status of nuclear power plants, but equipment failures and external interference will lead to missing operating monitoring data, which will reduce the quality of the data and thus reduce the accuracy of the subsequent analysis results. To solve this problem, this paper utilizes the fact that nuclear power plants have accumulated a large amount of operational data and researches the method of generating adversarial imputation network (GAIN)-based imputation method for missing values of nuclear power plants' operational data. The generator in the model estimates the missing values by learning the distribution of the true values, and the discriminator in the model discriminates which values are true and which are generated with the help of a hint matrix. The hint reveals partial information about the missing original samples to the discriminator, which the discriminator uses to focus its attention on the quality of the imputation of particular data values. Finally, a training set and a test set were constructed for comparative experiments on the PCTran simulation platform by simulating the operational data of the AP1000 as an example. The experimental results demonstrate that the investigated algorithm achieves lower root mean square error (RMSE), verifying the feasibility and accuracy of the method. © 2024 The authors and IOS Press.

关键词： Nuclear power plants

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：