检索结果-内蒙古大学图书馆

Robust Deep Transfer Learning Based Object Detection and Tracking Approach

Intelligent Automation & Soft Computing 2023年第3期35卷 3613-3626页

作者： C.Narmadha T.Kavitha R.Poonguzhali V.Hamsadhwani Ranjan walia Monia B.Jegajothi Electronics and Communication Engineering Periyar Maniammai Institute of Science&Technology(PMIST)Thanjavur613403India Department of Computer Science and Engineering Periyar Maniammai Institute of Science&TechnologyThanjavur613403India Department of Electrical and Electronics Engineering Periyar Maniammai Institute of Science and TechnologyThanjavur613403India Electrical Engineering Department Model Institute of Engineering&Technology(Autonomous)Jammu181122India Department of Computer Science and Engineering Shri Mata Vaishno Devi UniversityKatra182320India Center for Energy Research Chennai Institute of TechnologyChennai600069India

At present days,object detection and tracking concepts have gained more importance among researchers and business ***,deep learning(DL)approaches have been used for object tracking as it increases the perfor-mance and speed of the tracking *** paper presents a novel robust DL based object detection and tracking algorithm using Automated Image Anno-tation with ResNet based Faster regional convolutional neural network(R-CNN)named(AIA-FRCNN)*** AIA-RFRCNN method performs image anno-tation using a Discriminative Correlation Filter(DCF)with Channel and Spatial Reliability tracker(CSR)called DCF-CSRT *** AIA-RFRCNN model makes use of Faster RCNN as an object detector and tracker,which involves region proposal network(RPN)and Fast *** RPN is a full convolution network that concurrently predicts the bounding box and score of different *** RPN is a trained model used for the generation of the high-quality region proposals,which are utilized by Fast R-CNN for detection ***,Residual Network(ResNet 101)model is used as a shared convolutional neural network(CNN)for the generation of feature *** performance of the ResNet 101 model is further improved by the use of Adam optimizer,which tunes the hyperparameters namely learning rate,batch size,momentum,and weight ***,softmax layer is applied to classify the *** performance of the AIA-RFRCNN method has been assessed using a benchmark dataset and a detailed comparative analysis of the results takes *** outcome of the experiments indicated the superior characteristics of the AIA-RFRCNN model under diverse aspects.

关键词： Object detection tracking deep learning deep transfer learning image annotation

来源：评论

学校读者我要写书评

暂无评论

Learning-Based Joint Service Caching and Load Balancing for MEC Blockchain Networks

引用

China Communications 2023年第1期20卷 125-139页

作者： Wenqian Zhang Wenya Fan Guanglin Zhang Shiwen Mao College of Information Science and Technology Donghua UniversityShanghai 201620China Engineering Research Center of Digitized Textile and Apparel Technology Ministry of EducationShanghai 201620China Department of Electrical and Computer Engineering Auburn UniversityAuburnAL 36849-5201USA

Integrating the blockchain technology into mobile-edge computing(MEC)networks with multiple cooperative MEC servers(MECS)providing a promising solution to improving resource utilization,and helping establish a secure reward mechanism that can facilitate load balancing among *** addition,intelligent management of service caching and load balancing can improve the network utility in MEC blockchain networks with multiple types of *** this paper,we investigate a learningbased joint service caching and load balancing policy for optimizing the communication and computation resources allocation,so as to improve the resource utilization of MEC blockchain *** formulate the problem as a challenging long-term network revenue maximization Markov decision process(MDP)*** address the highly dynamic and high dimension of system states,we design a joint service caching and load balancing algorithm based on the double-dueling Deep Q network(DQN)*** simulation results validate the feasibility and superior performance of our proposed algorithm over several baseline schemes.

关键词： cooperative mobile-edge computing blockchain workload offloading service caching load balancing deep reinforcement learning(DRL)

来源：评论

学校读者我要写书评

暂无评论

ElectrodeNet - A Deep-Learning-Based Sound Coding Strategy for Cochlear Implants

引用

IEEE Transactions on Cognitive and Developmental Systems 2024年第1期16卷 346-357页

作者： Huang, Enoch Hsin-Ho Chao, Rong Tsao, Yu Wu, Chao-Min Department of Electrical Engineering National Central University Taoyuan320317 Taiwan Research Center for Information Technology Innovation Academia Sinica Taipei115201 Taiwan Department of Computer Science and Information Engineering National Cheng Kung University Tainan701401 Taiwan Department of Electrical Engineering Chung Yuan Christian University Taoyuan320314 Taiwan

ElectrodeNet, a deep-learning-based sound coding strategy for the cochlear implant (CI), is proposed to emulate the advanced combination encoder (ACE) strategy by replacing the conventional envelope detection using various artificial neural networks. The extended ElectrodeNet-CS strategy further incorporates the channel selection (CS). Network models of deep neural network (DNN), convolutional neural network (CNN), and long short-term memory (LSTM) were trained using the fast Fourier transformed bins and channel envelopes obtained from the processing of clean speech by the ACE strategy. Objective speech understanding using short-time objective intelligibility (STOI) and normalized covariance metric (NCM) was estimated for ElectrodeNet using CI simulations. Sentence recognition tests for vocoded Mandarin speech were conducted with normal-hearing listeners. DNN, CNN, and LSTM-based ElectrodeNets exhibited strong correlations to ACE in objective and subjective scores using mean squared error (MSE), linear correlation coefficient (LCC), and Spearman's rank correlation coefficient (SRCC). The ElectrodeNet-CS strategy was capable of producing N-of-M compatible electrode patterns using a modified DNN network to embed maxima selection, and to perform in similar or even slightly higher average in STOI and sentence recognition compared to ACE. The methods and findings demonstrated the feasibility and potential of using deep learning in the CI coding strategy. © 2023 IEEE.

关键词： Cochlear implants

来源：评论

学校读者我要写书评

暂无评论

Gradual Domain Adaptation via Manifold-Constrained Distributionally Robust Optimization 38

Gradual Domain Adaptation via Manifold-Constrained Distribut...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Saberi, Amirhossein Najafi, Amir Behjati, Amin Emrani, Ala Zolfi, Yasaman Shadrooy, Mahdi Motahari, Abolfazl Khalaj, Babak H. Department of Electrical Engineering Sharif University of Technology Tehran Iran Department of Computer Engineering Sharif University of Technology Tehran Iran Sharif Center for Information Systems and Data Science Sharif University of Technology Tehran Iran

The aim of this paper is to address the challenge of gradual domain adaptation within a class of manifold-constrained data distributions. In particular, we consider a sequence of T ≥ 2 data distributions P1, ..., PT undergoing a gradual shift, where each pair of consecutive measures Pi, Pi+1 are close to each other in Wasserstein distance. We have a supervised dataset of size n sampled from P0, while for the subsequent distributions in the sequence, only unlabeled i.i.d. samples are available. Moreover, we assume that all distributions exhibit a known favorable attribute, such as (but not limited to) having intra-class soft/hard margins. In this context, we propose a methodology rooted in Distributionally Robust Optimization (DRO) with an adaptive Wasserstein radius. We theoretically show that this method guarantees the classification error across all Pis can be suitably bounded. Our bounds rely on a newly introduced compatibility measure, which fully characterizes the error propagation dynamics along the sequence. Specifically, for inadequately constrained distributions, the error can exponentially escalate as we progress through the gradual shifts. Conversely, for appropriately constrained distributions, the error can be demonstrated to be linear or even entirely eradicated. We have substantiated our theoretical findings through several experimental results. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Equipment Parameter Investigation for a Proposed Power Conductor Hot Spot Identification system 10

Equipment Parameter Investigation for a Proposed Power Condu...

引用

10th International Conference on Modern Power Systems, MPS 2023

作者： Polycarpou, Alexis Komodromos, Michael Kalli, Kyriacos Argyrou, Maria C. Ioannou, Andreas Computer Engineering and Informatics Frederick University Frederick Research Center Deparment of Electrical Nicosia Cyprus Cyprus University of Technology PhOSLab Department of Electrical Engineering/Computer Engineering and Informatics Limassol Cyprus

ISBN: (纸本)9798350326826

Reliability, efficiency and continuity of power energy supplied is an area which receives increasing attention as the main infrastructure of power transmission and distribution systems in many countries is ageing. Hot spot related cable failures and power interruptions have a big financial impact on the power provider. The operational parameters of a method to identify in real time hot spots on cables is investigated in this paper, which can result to timely fault prevention. This can be achieved through exploiting the optical sensing capabilities of the existing optical fibre on-grid network. The resulting parameters aim to be advantageous for the system as they increase accuracy and will enable the Remote, automated, continuous, and real time monitoring of the grid infrastructure integrity. © 2023 IEEE.

关键词： Cables

来源：评论

学校读者我要写书评

暂无评论

Efficient Object Detection Acceleration Methods for Autonomous-driving Embedded Platforms

引用

IEIE Transactions on Smart Processing and Computing 2022年第4期11卷 255-261页

作者： Choi, Jiwoong Chun, Dayoung Lee, Hyuk-Jae Kim, Hyun Department of Electrical and Computer Engineering Seoul National University Seoul08826 Korea Republic of Research Center for Electrical and Information Technology Department of Electrical and Information Engineering Seoul National University of Science and Technology Seoul01811 Korea Republic of

Object detection in autonomous vehicles is typically operated in an embedded system to reduce power consumption. The use of an object detection algorithm with high accuracy and real-time detection speed in the embedded systems is essential for ensuring safe driving. This study proposes a parallel processing method for GPU and CPU operations to enhance the detection speed of the model. In addition, this study proposes data augmentation and image resize techniques that consider the camera input size of autonomous driving, which increases the accuracy significantly while improving the detection speed. The application of these proposed schemes to a baseline algorithm, tiny Gaussian YOLOv3, improves the mean average precision by 1.14 percent points (pp) for the Berkeley Deep Drive (BDD) dataset and 1.34 pp for the KITTI dataset compared to the baseline. Furthermore, in the NVIDIA Jetson AGX Xavier, which is an embedded platform for autonomous driving, the proposed algorithm improves the detection speed by 22.54 % for the BDD, and 24.67 % for the KITTI compared to the baseline, thereby enabling high-speed real-time detection on both datasets. © 2022 Institute of Electronics and Information Engineers. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Distributed Average Integral Control Based Energy Management Model

Distributed Average Integral Control Based Energy Management...

引用

2023 IEEE International Conference on Energy Technologies for Future Grids, ETFG 2023

作者： Shaukat, Neelofar Khan, Bilal Qureshi, Muhamamd Bilal Jawad, Muhammad Shabbir, Noman Daniel, Kamran Abbottabad Pakistan Lahore Pakistan FinEST Center for Smart Cities Department of Electrical and Computer Engineering Estonia Tallinn University of Technology Department of Electrical Power Engineering & Mechatronics Tallinn Estonia Nova University Department of Electrical & Computer Engineering Lisbon Portugal

ISBN: (纸本)9781665471640

In this research study, we proposed a Distributed Average Integral (DAI) control-based Energy Management Model (EMM) to achieve the economic load dispatch and consensus within a distributed energy system. The proposed model incorporates the Laplacian graph theory for establishing communication among energy districts (acting as agents/nodes) and optimizing the power distribution while satisfying the load demand. In order to evaluate the effectiveness of the proposed model, simulations are performed using MATLAB. The simulation results illustrate that the DAI control mechanism ensures the optimized power distribution leading to enhanced resource utilization while effectively achieving economic consensus among energy districts. However, in this study ideal case scenario is assumed. Therefore, this research study may highlight the potential for future investigations on utilizing the DAI control for EMM in practical scenarios with real-world challenges such as system complexities, losses, and communication delays. © 2023 IEEE.

关键词： Distributed Average Integral Control Energy Districts Graph Theory Power Consensus

来源：评论

学校读者我要写书评

暂无评论

Learning Compositional, Time-Varying Neural Barrier Contracts 27

Learning Compositional, Time-Varying Neural Barrier Contract...

引用

27th European Conference on Artificial Intelligence, ECAI 2024

作者： Low, Matthew Wang, Timothy E. Nuzzo, Pierluigi Department of Electrical and Computer Engineering University of Southern California Los AngelesCA United States RTX Technology Research Center BerkeleyCA United States Department of Electrical Engineering and Computer Sciences University of California BerkeleyCA United States

ISBN: (纸本)9781643685489

Certificate functions can be used to efficiently capture and prove various properties of a system or controller. Control barrier functions (CBFs) are certificate functions that define a region of forward-invariance, which makes them a natural choice to enforce a notion of "safety" for a system. However, synthesizing CBFs, especially in the case of complex systems with learning-enabled components, remains a challenge. Recent work achieves promising results by leveraging neural networks as function approximators to learn CBFs. However, a single CBF may not exist or be difficult to obtain for complex hybrid systems. To overcome this difficulty, this paper presents a framework to simultaneously learn simpler, time-varying control barrier functions (TV-CBFs) that are composable. We embed these neural certificates in a compositional framework based on assume-guarantee contracts. The resulting neural barrier contracts can then be combined by leveraging the rigorous contract algebra. Learning multiple, composable CBFs empowers the verification process (1) by simplifying the verification of complex systems via decomposition and (2) by broadening the expressivity in capturing complex systems and controllers. We illustrate the effectiveness of our approach on the verification of an aircraft's automatic landing system and a quadrotor navigating an indoor environment. © 2024 The Authors.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Crack Detection for Non-Metallic Material Using Split Ring Resonator-based Chipless RFID Sensor 20

Crack Detection for Non-Metallic Material Using Split Ring R...

引用

20th International Conference on electrical engineering/Electronics, computer, Telecommunications and Information technology, ECTI-CON 2023

作者： Nongpromma, Sirisak Janin, Suwan Kaemarungsi, Kamol Boonpoonga, Akkarat Torrungrueng, Danai Bannawat, Lakkhana King Mongkut's University of Technology North Bangkok Faculty of Engineering Department of Electrical and Computer Engineering Bangkok Thailand Rajamangala University of Technology Lanna Lampang Faculty of Engineering Department of Engineering and Technology Lampang52000 Thailand National Electronics and Computer Technology Center National Science and Technology Development Agency Pathumthani Thailand Department of Electrical and Computer Engineering Bangkok Thailand Department of Teacher Training in Electrical Engineering Bangkok Thailand Rajamangala University of Technology Rattanakosin Faculty of Engineering Department of Computer Engineering Nakhon Pathom Thailand

ISBN: (纸本)9798350310467

The study presented a novel chipless RFID sensor designed for the detection of cracks in non-metallic materials. The sensor utilized square coplanar split-ring resonators (SRR) and was mounted on the material under test (MUT), including materials such as Teflon (PTFE), Polycarbonate, and Kapton (Polyimide), for detecting cracks. To assess the effectiveness of the sensor in detecting cracks, simulation experiments were conducted using the CST Microwave Studio software, which analyzed shifts in the resonant frequency. The resulting RCS spectra demonstrated a clear correlation between the size of the crack and the resonance frequencies of the chipless RFID sensor, confirming the sensor's ability to detect cracks in nonmetallic materials. © 2023 IEEE.

关键词： Nonmetallic materials

来源：评论

学校读者我要写书评

暂无评论

DRL-Driven Localization With UAV in Near-Field Communications

引用

IEEE Internet of Things Journal 2025年第13期12卷 22587-22598页

作者： Khan, Muhammad Fawad Peng, Limei Ho, Pin-Han Chen, Yuguang Dong, FangJie Kyungpook National University School of Computer Science and Engineering Daegu41566 Korea Republic of University of Electronic Science and Technology of China Shenzhen Institute for Advanced Study Guangdong China University of Waterloo Department of Electrical and Computer Engineering ON Canada Shenzhen Unicom Innovation Business Capability Center China National Health Commission's Center for Statistics and Information China

In this paper, we propose a Deep Reinforcement Learning (DRL)-based Multi-Point Localization Scheme (MLS) to efficiently localize IoT devices using a single Unmanned Aerial Vehicle (UAV) equipped with a large-scale multi-antenna configuration in near-field communication (NFC). By utilizing the spherical wave-based near-field steering vector, the multi-antenna array on the UAV captures both the Angle of Arrival (AoA) and Received Signal Strength Indicator (RSSI) measurements from IoT devices to estimate their locations relative to the position of the UAV. This approach eliminates the need for multiple hovering points required by a single-antenna UAV (SA-UAV) or the deployment of multiple SA-UAVs. To enhance localization accuracy, key hovering points for the multi-antenna UAV (MA-UAV) are strategically selected, with weights assigned based on signal strength to prioritize stronger and more reliable signals. Furthermore, DRL dynamically adjusts the position of the MA-UAV to optimize the trade-off between localization accuracy and energy consumption. Extensive simulations conducted across rural, urban, and dense urban scenarios demonstrate that the proposed DRL-based MLS significantly improves localization accuracy while reducing the energy consumption of the UAV. © 2025 IEEE.

关键词： Deep reinforcement learning

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：