检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Rosin, David Kässinger, Johannes Yu, Xingyao Avci, Okan Bleiler, Christian Röhrle, Oliver Institute for Modelling and Simulation of Biomechanical Systems University of Stuttgart Germany Institute for Parallel and Distributed Systems University of Stuttgart Germany Visualization Research Center VISUS University of Stuttgart Germany Biomechatronic Systems Fraunhofer IPA Stuttgart Germany University of Stuttgart Germany

This paper presents a novel neural network architecture for the purpose of pervasive visualisation of a 3D human upper limb musculoskeletal system model. Bringing simulation capabilities to resource-poor systems like mobile devices is of growing interest across many research fields, to widen applicability of methods and results. Until recently, this goal was thought to be out of reach for realistic continuum-mechanical simulations of musculoskeletal systems, due to prohibitive computational cost. Within this work we use a sparse grid surrogate to capture the surface deformation of the m. biceps brachii in order to train a deep learning model, used for real-time visualisation of the same muscle. Both these surrogate models take 5 muscle activation levels as input and output Cartesian coordinate vectors for each mesh node on the muscle’s surface. Thus, the neural network architecture features a significantly lower input than output dimension. 5 muscle activation levels were sufficient to achieve an average error of 0.97 ± 0.16 mm, or 0.57 ± 0.10 % for the 2809 mesh node positions of the biceps. The model achieved evaluation times of 9.88 ms per predicted deformation state on CPU only and 3.48 ms with GPU-support, leading to theoretical frame rates of 101 fps and 287 fps respectively. Deep learning surrogates thus provide a way to make continuum-mechanical simulations accessible for visual real-time applications. © 2023, CC BY-NC-ND.

关键词： Muscle

来源：评论

学校读者我要写书评

暂无评论

Efficient All-reduce for distributed DNN Training in Optical Interconnect System

arXiv

引用

arXiv 2022年

作者： Dai, Fei Chen, Yawen Huang, Zhiyi Zhang, Haibo Zhang, Fangfang University of Otago Dunedin New Zealand Qilu University of Technology Jinan China

All-reduce is the crucial communication primitive to reduce model parameters in distributed Deep Neural Networks (DNN) training. Most existing all-reduce algorithms are designed for traditional electrical interconnect systems, which cannot meet the communication requirements for distributed training of large DNNs due to the low data bandwidth of the electrical interconnect systems. One of the promising alternatives for electrical interconnect is optical interconnect, which can provide high bandwidth, low transmission delay, and low power cost. We propose an efficient scheme called Wrht (Wavelength Reused Hierarchical Tree) for implementing all-reduce operation in optical interconnect systems. Wrht can take advantage of WDM (Wavelength Division Multiplexing) to reduce the communication time of distributed data-parallel DNN training. Simulations using real DNN models show that, compared to all-reduce algorithms in the electrical and optical network systems, our approach reduces communication time by 75.76% and 91.86%, respectively. © 2022, CC BY-NC-ND.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Labeling Sentences with Symbolic and Deictic Gestures via Semantic Similarity

Labeling Sentences with Symbolic and Deictic Gestures via Se...

引用

IEEE International workshop on Robot and Human Communication (ROMAN)

作者： Ariel Gjaci Carmine Tommaso Recchiuto Antonio Sgorbissa University of Genova Genova Italy

ISBN: (数字)9798350375022

ISBN: (纸本)9798350375039

Co-speech gesture generation on artificial agents has gained attention recently, mainly when it is based on data-driven models. However, end-to-end methods often fail to generate co-speech gestures related to semantics with specific forms, i.e., Symbolic and Deictic gestures. In this work, we identify which words in a sentence are contextually related to Symbolic and Deictic gestures. Firstly, we appropriately chose 12 gestures recognized by people from the Italian culture, which different humanoid robots can reproduce. Then, we implemented two rule-based algorithms to label sentences with Symbolic and Deictic gestures. The rules depend on the semantic similarity scores computed with the RoBerta model between sentences that heuristically represent gestures and sub-sentences inside an objective sentence that artificial agents have to pronounce. We also implemented a baseline algorithm that assigns gestures without computing similarity scores. Finally, to validate the results, we asked 30 persons to label a set of sentences with Deictic and Symbolic gestures through a Graphical User Interface (GUI), and we compared the labels with the ones produced by our algorithms. For this scope, we computed Average Precision (AP) and Intersection Over Union (IOU) scores, and we evaluated the Average Computational time (ACT). Our results show that semantic similarity scores are useful for finding Symbolic and Deictic gestures in utterances.

关键词： Measurement Scalability Semantics Training data Humanoid robots parallel processing real-time systems Hybrid power systems Labeling Graphical user interfaces

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of Data-Centric Networks Based on parallel and distributed Architectures for Machine Intelligence Research

引用

parallel PROCESSING LETTERS 2023年第3期33卷 2340006-2340006页

作者： Xie, Linjiang Hang, Feilu Guo, Wei Zhang, Zhenhong Li, Hanruo Power Grid Co Ltd Informat Ctr Yunnan Kunming 650000 Yunnan Peoples R China

Maintaining computer network security has long been an essential component of computer administration. Network security has become essential to companies' safety and steady development in real-time implementation for diverse physical domains employing computers. Establishing a specific physical domain has consistently included a priority job of improving the safety of computer systems management. Study results on the practical implementation of machine intelligence in network security maintenance. Specifically, this paper examines and offers security management techniques for computer network data protection to establish an all-encompassing security shield for networked computers, with the ultimate goal of improving the safety and reliability of the net-worked computers used in commercial settings. Furthermore, the comprehensive research analysis determines an architectural model for machine intelligence-based network security maintenance (MI-NSM), implying an intrusion detection scheme with a novel neural network system. Labelling important properties of objects or data points and searching for commonalities allows the auto-mobile's AI to distinguish between a human, the street, another car, and the sky. The simulation evaluation is performed using the NS2 simulator and observes the security maintenance efficiency over the current security solutions.

关键词： Machine intelligence network security defence intrusion detection neural network

来源：评论

学校读者我要写书评

暂无评论

Evaluation of the impact on industrial applications of NTP used by IoT devices

Evaluation of the impact on industrial applications of NTP u...

引用

IEEE International workshop on Metrology for Industry 4.0 and IoT (METROIND4.O and IoT)

作者： Ferrari, Paolo Bellagente, Paolo Depari, Alessandro Flammini, Alessandra Pasetti, Marco Rinaldi, Stefano Sisinni, Emiliano Univ Brescia Dept Informat Engn Brescia Italy

ISBN: (纸本)9781728148922

The requirements of distributed systems implemented using IoT paradigms include the sharing of a common time reference. When the applications are intended for industrial scenarios, the desired time synchronization uncertainty decreases, due to the real-time needs of this kind of systems. Currently, the most diffused way to obtain a time reference for an Internet connected device is to use NTP (Network time Protocol) for querying time servers. This paper is focused on industrial IoT applications and it is aimed to evaluate the impact that NTP synchronization performance and NTP throughput may have on the real-time host network. Analytical formulas for estimating the NTP impact are introduced. The work is complemented by the presentation of a use case that shows the use of NTP for the synchronization of a large industrial plant at field level. The results highlight the threat to the real-time behavior of the automation system caused by the uncontrolled peaks of traffic due to NTP. Last, some design considerations for mitigation of the impact are discussed.

关键词： Industrial automation distributed measurement system network performance distributed clock

来源：评论

学校读者我要写书评

暂无评论

Initial Results on the Effectiveness of distributed Edge Detection for Large Image Data Sets

Initial Results on the Effectiveness of Distributed Edge Det...

引用

IEEE Southeastcon

作者： Ivanescu Constantin Renato Mocanu Mihai University of Craiova Craiova Romania

Image data is expanding rapidly, along with technology development, so efficient solutions must be considered to achieve high, real-time performance in the case of processing large image datasets. parallel processing is increasingly used as an attractive alternative to improve the performance, when using existing distributed architectures but also for sequential commodity computers. It can provide speedup, efficiency, reliability, incremental growth, and flexibility. We present such an alternative and stress the effectiveness of the methods to accelerate computations on a small cluster of PCs compared to a single CPU. Our paper is focused on applying edge detection on large image data sets, as a fundamental and challenging task in image processing and computer vision. Five different techniques, mainly Sobel, Prewitt, LoG, Canny, and Roberts, are compared in a simple experimental setup that includes the OpenCV library functions for image pixels manipulation. Gaussian blur is used to reduce high-frequency components to manage the noise that edge detection is impacted by. Overall, this work is part of a more extensive investigation of image segmentation methods on large image datasets, but the results presented are relevant and show the effectiveness of our approach.

关键词： Computers Image segmentation Image edge detection distributed databases parallel processing real-time systems Libraries

来源：评论

学校读者我要写书评

暂无评论

Combination of Stochastic State Estimation with Online Identification of the Open-Circuit Voltage of Lithium-Ion Batteries 1

Combination of Stochastic State Estimation with Online Ident...

引用

1st IFAC workshop on Control of Complex systems (COSY)

作者： Lahme, Marit Rauh, Andreas Carl von Ossietzky Univ Oldenburg Dept Comp Sci Grp Distributed Control Interconnected Syst D-26111 Oldenburg Germany

Typical equivalent circuit models for Lithium-ion batteries are represented in the form of the so-called Thevenin form. These models comprise a state of charge (SOC) controlled voltage source as well as lumped elements such as series resistances as well as at least one RC subnetwork to describe dynamic effects by the terminal current and voltage as input and output of the battery. In classical state estimation approaches, these characteristics are assumed to be identified beforehand, while aging of battery cells is related to a loss of the total capacity, changes in the charging/discharging efficiency, an increase of the Ohmic cell resistance, and/or changes of the time constants of the above-mentioned RC networks. Such variations can be estimated by means of additional parameters that are included in the system's state vector. However, the typically applied approaches do not allow for a direct identification of the nonlinear dependencies of the circuit elements on the SOC and other influence factors such as currents or the cell temperature. Therefore, this paper proposes a real-time capable, two-stage identification of these dependencies on the example of the open-circuit voltage by means of an Unscented Kalman Filter (UKF) approach. In the first stage, the state variables of the dynamic system model are estimated together with a lumped disturbance term (an additive correction of the open-circuit voltage). In the second stage, this disturbance term is used to correct the a-priori knowledge of the open-circuit voltage and to simultaneously express the identification quality in terms of covariances for the coefficients of its piecewise polynomial approximation in each temporal discretization step. The efficiency of the proposed methodology is demonstrated in simulations for an experimentally validated system model. Copyright (c) 2022 The Authors. This is an open access article under the CC BY-NC-ND license (https://***/licenses/by-nc-nd/4.0/)

关键词： Online identification stochastic state estimation uncertainty quantification Unscented Kalman Filter Lithium-ion batteries

来源：评论

学校读者我要写书评

暂无评论

An Investigation of Series and parallel Configurations for Hybrid Power Amplifiers

An Investigation of Series and Parallel Configurations for H...

引用

作者： Matiuzzi Kunzler, Luccas Concordia University

学位级别：博士

Power Hardware-in-the-Loop (PHIL) is becoming increasingly popular for compartmentalized testing of electric power equipment in several areas such as in electric drive systems and distributed power generation systems. The fundamental idea of PHIL is to create flexible conditions for Devices under Test (DUT) to be properly assessed in real time and dynamic conditions with their rated power levels. Connected to the DUT is the Power Amplifier (PA), which is responsible for increasing the voltage and current levels, given from the real-time Simulator (RTS). The DUT is a physical equipment and high-complexity models are used to control the PAs to emulate necessary conditions for the DUT to be evaluated. One of the main benefits of PHIL is that it can provide a platform for conducting a number of severe tests without risking damaging the equipment that is being emulated, while testing the actual response of the DUT. It can also help with the preliminary design and performance assessment of new types of machines, drivers and controllers, thus significantly reducing the time to market of new equipment. The flexibility of PHIL is also one of its main assets, since the combination of the RTS and the PA can be used for various applications only by changing the model and/or parameters of the emulated element. This thesis will evaluate the main architectures, control strategies and PHIL applications of PAs. Linear Power Amplifiers (LPA) provide an overall great performance due to its high bandwidth but are expensive, mostly at increased power ratings. For high PAs with fast dynamic response and reduced waveform distortion, the Hybrid Power Amplifier (HPA) configuration provides a good cost-performance compromise. HPAs are built essentially with the association of a low-cost Switch Mode Power Amplifier (SMPA) and an LPA. The first configuration to be investigated is the series connected HPA intended for high voltage systems. The SMPA consists of a Cascaded H-Bridge Multilevel (CH

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fiber Optic Monitoring and Forecasting of Reservoir Landslides 1

引用

8th Civil Structural Health Monitoring workshop (CSHM)

作者： Zhu, Hong-Hu Shi, Bin Zhang, Lei Ye, Xiao Nanjing Univ Sch Earth Sci & Engn Nanjing 210023 Peoples R China

ISBN: (数字)9783030742584

ISBN: (纸本)9783030742607;9783030742584;9783030742577

Landslide hazards triggered by atmospheric precipitation and periodic water level fluctuations frequently occur in reservoir areas all over the world, resulting in severe fatalities and loss of properties. For the purpose of detecting geological disasters ahead of time, a series of advance monitoring techniques have been developed and applied to engineering practices. Among them, the distributed Fiber Optic Sensing (DFOS) technologies have shown their great potential for monitoring the tempo-spatial distribution of multi-fields (displacement, stress, temperature, seepage, etc.) of reservoir landslides. In this paper, the working principles, characteristics and applicability and these technologies are briefly introduced. In particularly, several representative DFOS-based monitoring instruments and equipment are elaborated, which have been successfully used in several landslides in Three Gorges Reservoir (TGR) area in China for long-term in situ observation and comprehensive evaluation of relevant geohazards. The monitoring results of Majiagou landslide show that the sliding surfaces, deformation pattern, and groundwater evolution can be sensitively captured based on the DFOS measurements. Seasonal rainfall and water level fluctuations are found to be the two main triggering factors of the landslide. Meanwhile, long time-series monitoring data shed a bright light on real-time deformation prediction and stability analyses of reservoir landslides. Finally, the challenges and future trends of applying the DFOS systems to landslide monitoring are summarized.

关键词： distributed fiber optic sensing (DFOS) Field monitoring Reservoir landslides Three gorges reservoir (TGR)

来源：评论

学校读者我要写书评

暂无评论

Predictable remote invocations for distributed stream processing

引用

FUTURE GENERATION COMPUTER systems-THE INTERNATIONAL JOURNAL OF ESCIENCE 2020年 107卷 716-729页

作者： Basanta-Val, P. Fernandez-Garcia, N. Sanchez-Fernandez, L. Univ Carlos III Madrid UC3M BS Inst Financial Big Data IFiBiD Dept Ingn Telemat Madrid 28911 Spain Ctr Univ Defensa Escuela Naval Mil Plaza Espana S-N Marin 36920 Spain

Typical infrastructure for big-data includes multiple machines with data accessed remotely with request-response patterns from different remote locations. Currently, most of the state-of-the-art remote invocation techniques are focused on models for distributed interactions, which have not explored the advantages given by parallel computing, such as those offered to run on distributed stream processors. In this context, the article is focused on the definition of a predictable remote procedure call (RPC) able to take advantage from the distributed stream processing technology. Potentially, this type of infrastructure enables efficient parallel computations, reducing the effective response-time of end-to-end invocations linearly with the number of resources assigned to the system, as the data increases. The article describes a predictable model which defines maximum response-times for different types of applications described in the context of Apache Storm. Evaluation also offers clues on the performance one may expect from this type of infrastructure. (c) 2017 Elsevier B.V. All rights reserved.

关键词： Big-data distributed stream processing Predictable infrastructure Apache storm time-critical big-data Efficient big-data

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：