检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Przewlocka-Rus, Dominika Kryjak, Tomasz Embedded Vision Systems Group Computer Vision Laboratory Department of Automatic Control and Robotics AGH University of Science and Technology Krakow Poland

Deep neural networks virtually dominate the domain of most modern vision systems, providing high performance at a cost of increased computational complexity. Since for those systems it is often required to operate both in real-time and with minimal energy consumption (e.g., for wearable devices or autonomous vehicles, edge Internet of Things (IoT), sensor networks), various network optimisation techniques are used, e.g., quantisation, pruning, or dedicated lightweight architectures. Due to the logarithmic distribution of weights in neural network layers, a method providing high performance with significant reduction in computational precision (for 4-bit weights and less) is the Power-of-Two (PoT) quantisation (and therefore also with a logarithmic distribution). This method introduces additional possibilities of replacing the typical for neural networks Multiply and ACcumulate (MAC – performing, e.g., convolution operations) units, with more energy-efficient Bitshift and ACcumulate (BAC). In this paper, we show that a hardware neural network accelerator with PoT weights implemented on the Zynq UltraScale + MPSoC ZCU104 SoC FPGA can be at least 1.4x more energy efficient than the uniform quantisation version. To further reduce the actual power requirement by omitting part of the computation for zero weights, we also propose a new pruning method adapted to logarithmic quantisation. © 2022, CC BY.

关键词： System-on-chip

来源：评论

学校读者我要写书评

暂无评论

Community Search Based on Containment control of Multi-Agent System with Opinion Leaders

Community Search Based on Containment Control of Multi-Agent...

引用

第43届中国控制会议

作者： Siyang Liu Tong Li Jinliang Shao Wei Xing Zheng School of Automation Engineering University of Electronic Science and Technology of China Laboratory of Electromagnetic Space Cognition and Intelligent Control School of Computer Data and Mathematical SciencesWestern Sydney University

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

Community search is the problem of identifying the community in which a given node *** from traditional community search methods based on specific topological structures,this paper proposes a community search algorithm based on containment control of multi-agent system by studying the influence of opinion leaders with competitive relationships in community ***,the Breadth-First Search method is used to locally sample the ***,containment control of multi-agent system is applied to obtain the final state of each node for local community ***,grounded in practical considerations,an effective leadership transfer mechanism is proposed to update the results and enhance search *** experiments on real-world networks and synthetic networks in comparison with three local community detection algorithms,the effectiveness and rationality of our algorithm are demonstrated.

关键词： Community Search Opinion Leaders Multi-Agent System Containment control

来源：评论

学校读者我要写书评

暂无评论

Distributed Traffic Signal control of Interconnected Intersections: A Two-Lane Traffic Network Model

arXiv

引用

arXiv 2024年

作者： Ru, Xinfeng Xia, Weiguo Bai, Ting Key Laboratory of Intelligent Control and Optimization for Industrial Equipment Dalian University of Technology Dalian116024 China School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm10044 Sweden

Practical and accurate traffic models play an important role in capturing real traffic dynamics and then in achieving effective control performance. This paper studies traffic signal control in a traffic network with multiple interconnected intersections, where the target is to balance the vehicle density on each lane by controlling the green times of each phase at every intersection. Different from traditional road-based modeling schemes, a two-lane intersection model is first proposed to model the flow propagation in a more accurate way. A distributed model predictive control (MPC) method is then presented to assign the green times. To enable the real-time feasibility of the proposed approach, the alternating direction method of multipliers (ADMM) is incorporated with the distributed MPC scheme for solving the problem. Finally, the simulation studies performed in VISSIM for a six-intersection traffic network in Dalian, China, show the effectiveness and characteristics of the proposed method. Copyright © 2024, The Authors. All rights reserved.

关键词： Model predictive control

来源：评论

学校读者我要写书评

暂无评论

A Novel Approach for Trajectory Partition Privacy in Location-Based Services

A Novel Approach for Trajectory Partition Privacy in Locatio...

引用

IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)

作者： Chundong Wang Yongxin Zhao School of Cyber Security and Computer National Engineering Laboratory for Computer Virus Prevention and Control Technology Tianjin University of Technology Tianjin China School of Computer Science and Engineering Engineering Research Center of Ministry of Education for Learning Intelligent Systems Tianjin University of Technology Tianjin China

ISBN: (数字)9798350381993

ISBN: (纸本)9798350382006

Location-based service (LBS) devices have enhanced daily life convenience but raised privacy concerns regarding user location data. For this, this research presents a novel approach to safeguard trajectory privacy through Similarity-based Trajectory Partitioning (STSM). The process begins by dividing trajectory equivalence classes based on distinct timestamps. Trajectory similarity is then assessed within each class using metrics such as Frechet distance, trajectory direction, and speed. Subsequently, a trajectory graph is constructed for each class, and the Dijkstra algorithm is applied to partition the graph using the k-subgraph method, effectively converting trajectory partitioning into graph partitioning. To meet publication requirements, Laplace noise is added to sensitive data linked to nodes and edges within the subgraph. Comparative experiments on real-world data confirm the efficacy, rationality, data availability, and privacy protection capabilities of the proposed trajectory partitioning algorithm, underscoring its superiority over alternative methods.

关键词： Weight measurement Privacy Differential privacy Conferences Noise Trajectory Partitioning algorithms

来源：评论

学校读者我要写书评

暂无评论

Goods Anomaly Detection Based on Multi-Branch Feature Extraction and Foreground Mining 14

Goods Anomaly Detection Based on Multi-Branch Feature Extrac...

引用

14th International Conference on Information technology in Medicine and Education, ITME 2024

作者： Jia, Zhiqian Li, Wei Li, Zuoyong Mu, Zhipeng Lin, Lihui Teng, Shenghua Shandong University of Science and Technology College of Electronic and Information Engineering Qingdao 266590 China Xiamen Air Traffic Management Station of Caac Xiamen 361006 China College of Computer and Data Science Minjiang University Fujian Provincial Key Laboratory of Information Processing and Intelligent Control Fuzhou 350121 China School of Mathematics and Computer Science Wuyi University Fujian Key Laboratory of Big Data Application and Intellectualization for Tea Industry Fujian 354300 China

ISBN: (纸本)9798350356670

In recent years, significant advancements have been made in the field of anomaly detection. Expanding anomaly detection tasks to new scenarios has become a development avenue. Anomaly detection for the outer packaging of supermarket goods is a challenging task due to the diversity of goods' appearances, which results in some detection methods suffering from limited sensitivity to anomalies and inadequate target mining. To address these issues, we propose a novel unsupervised anomaly detection method based on multi-branch feature extraction and foreground cosine structure mining strategy. Firstly, we use anomaly synthesis technique to enhance the model's perception of anomalies and construct more precise decision boundaries in the feature space. Secondly, a multi-branch feature extractor extracts and integrates multi-level and multi-scale features, providing rich texture and semantic information to improve the model's adaptability in real-world scenarios. Finally, the foreground cosine loss focuses on target mining while supplementing local structural information, further enhancing the model's ability to detect subtle anomalies. To validate the effectiveness of the proposed method, extensive experiments were conducted on the GoodsAD dataset, sampled from supermarket scenarios. Experimental results show that this method achieves significant performance in image-level detection and pixel-level localization. © 2024 IEEE.

关键词： Anomaly Detection Foreground Mining Knowledge Distillation

来源：评论

学校读者我要写书评

暂无评论

Hardware-in-the-loop simulation of a UAV autonomous landing algorithm implemented in SoC FPGA

arXiv

引用

arXiv 2022年

作者： Szolc, Hubert Kryjak, Tomasz Embedded Vision Systems Group Computer Vision Laboratory Department of Automatic Control and Robotics AGH University of Science and Technology Krakow Poland

This paper presents a system for hardware-in-the-loop (HiL) simulation of unmanned aerial vehicle (UAV) control algorithms implemented on a heterogeneous SoC FPGA computing platforms. The AirSim simulator running on a PC and an Arty Z7 development board with a Zynq SoC chip from AMD Xilinx were used. Communication was carried out via a serial USB link. An application for autonomous landing on a specially marked landing strip was selected as a case study. A landing site detection algorithm was implemented on the Zynq SoC platform. This allowed processing a 1280 × 720 @ 60 fps video stream in real time. Performed tests showed that the system works correctly and there are no delays that could negatively affect the stability of the control. The proposed concept is characterised by relative simplicity and low implementation cost. At the same time, it can be applied to test various types of high-level perception and control algorithms for UAV implemented on embedded platforms. We provide the code developed on GitHub, which includes both Python scripts running on the PC and C code running on Arty Z7. © 2022, CC BY.

关键词： System-on-chip

来源：评论

学校读者我要写书评

暂无评论

Method for Evaluating the Performance of Aircraft Collaborative Decision-Making Based on Dynamic Bayesian Network

Method for Evaluating the Performance of Aircraft Collaborat...

引用

Electronics and Devices, Computational science (ICEDCS), International Conference on

作者： Wenbo Zhang Boxiao Yu Jiahui Tong Jing Li National Key Laboratory of Complex System Control and Intelligent Agent Cooperation Beijing Electro-Mechanical Engineering Institute Beijing China School of Computer Science and Technology Beijing Institute of Technology Beijing China

ISBN: (数字)9798331527624

ISBN: (纸本)9798331527631

This research addresses the challenge of assessing equipment decision-making effectiveness in environments characterized by uncertainty and multiple influencing factors. We propose a novel approach involving the creation and optimization of a dynamic Bayesian network-based model tailored for equipment decision-making processes. The model facilitates a systematic evaluation of decision-making effectiveness by incorporating various factors, including equipment performance, environmental conditions, and enemy defensive capabilities. A comprehensive assessment system has been developed to simulate the penetration capabilities of different equipment through enemy defenses. Moreover, this study identifies potential errors in the effectiveness evaluation model by analyzing its performance in typical scenarios, pinpointing areas for enhancement.

关键词： Accuracy Uncertainty Systematics Scientific computing Atmospheric modeling Computational modeling Decision making Data models Bayes methods Optimization

来源：评论

学校读者我要写书评

暂无评论

Star-Searcher: A Complete and Efficient Aerial System for Autonomous Target Search in Complex Unknown Environments

arXiv

引用

arXiv 2024年

作者： Luo, Yiming Zhuang, Zixuan Pan, Neng Feng, Chen Shen, Shaojie Gao, Fei Cheng, Hui Zhou, Boyu Sun Yat-Sen University China State Key Laboratory of Industrail Control Technology Institute of Cyber-Systems and Control Zhejiang University Hangzhou China Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology Hong Kong

This paper tackles the challenge of autonomous target search using unmanned aerial vehicles (UAVs) in complex unknown environments. To fill the gap in systematic approaches for this task, we introduce Star-Searcher, an aerial system featuring specialized sensor suites, mapping, and planning modules to optimize searching. Path planning challenges due to increased inspection requirements are addressed through a hierarchical planner with a visibility-based viewpoint clustering method. This simplifies planning by breaking it into global and local sub-problems, ensuring efficient global and local path coverage in real-time. Furthermore, our global path planning employs a history-aware mechanism to reduce motion inconsistency from frequent map changes, significantly enhancing search efficiency. We conduct comparisons with state-of-the-art methods in both simulation and the real world, demonstrating shorter flight paths, reduced time, and higher target search completeness. Our approach will be open-sourced for community benefit. Copyright © 2024, The Authors. All rights reserved.

关键词： Stars

来源：评论

学校读者我要写书评

暂无评论

Multi-View Landmark-Assisted UAV Swarm 6-DoF Pose Estimation Using Resonant Beam and VIO

引用

IEEE Internet of Things Journal 2025年

作者： Du, Siyuan Xu, Mengyuan Li, Yixuan Fang, Wen Liu, Qingwen Yi, Peng Hong, Yiguang Tongji University School of Computer Science and Technology Shanghai201804 China Tongji University College of Electronic and Information Engineering Shanghai201804 China Chinese Academy of Sciences Key Laboratory of Systems and Control Academy of Mathematics and Systems Science Beijing100190 China

As an essential aerial platform in Internet of Things (IoT) applications, UAV swarms require high-precision attitude estimation in GPS-limited and dynamic environments, which supports higher-level IoT functions such as smart logistics, disaster response, and environmental monitoring. However, most odometry-based pose estimation methods in dynamic scenarios without GPS encounter issues with cumulative errors over time. In this paper, we propose a method to reduce these cumulative errors by constraining the absolute positioning of Visual-Inertial Odometry (VIO) using relative poses obtained from multi-view landmark and Resonant-Beam (RBeam) sensors between UAVs. For synchronous moments, we estimate the 6 Degree-of-Freedom (DoF) relative pose using the Angle of Arrival and Time of Flight data from the RBeam, combined with nonlinear optimization. For asynchronous moments, a two-stage visual estimation method is introduced, combining multi-camera epipolar geometry for rotation recovery and depth reconstruction optimization for translation recovery, enabling the estimation of asynchronous relative poses. Finally, we design a global objective function based on a sliding window and factor graph, integrating RBeam, multi-view landmark, and VIO for absolute 6-DoF pose optimization of the UAV swarm. Simulation results demonstrate that optimizing with the addition of RBeam synchronous relative poses improves overall positioning accuracy by 32.94% compared to pure VIO. Incorporating both RBeam synchronous and visual asynchronous relative poses further enhances overall positioning accuracy by 37.65%. Additionally, the UAV’s attitude benefits from the rotational constraints provided by the RBeam, achieving over 30% improvement in the pitch and yaw directions. © 2014 IEEE.

关键词： Global positioning system

来源：评论

学校读者我要写书评

暂无评论

Energy Efficient Hardware Acceleration of Neural Networks with Power-of-Two Quantisation

TechRxiv

引用

TechRxiv 2022年

关键词： System-on-chip

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

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

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：