检索结果-内蒙古大学图书馆

IEEE Conference on Global Communications (GLOBECOM)

作者： Momin Abbas Koushik Kar Tianyi Chen Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute Troy NY USA

ISBN: (数字)9798350351255

ISBN: (纸本)9798350351262

Deep neural networks (DNNs) have made significant strides in tackling challenging tasks in wireless systems, especially when an accurate wireless model is not available. However, when available data is limited, traditional DNNs often yield subpar results due to underfitting. At the same time, large language models (LLMs) exemplified by GPT-3, have remarkably showcased their capabilities across a broad range of natural language processing *** whether and how LLMs can benefit challenging non-language tasks in wireless systems is *** this work, we propose to leverage the in-context learning ability (a.k.a. prompting) of LLMs to solve wireless tasks in the low data regime without any training or fine-tuning, unlike DNNs which require training. We further demonstrate that the performance of LLMs varies significantly when employed with different prompt templates. To solve this issue, we employ the latest LLM calibration methods. Our results reveal that using LLMs via ICL methods generally outperforms traditional DNNs on the symbol demodulation task and yields highly confident predictions when coupled with calibration techniques.

关键词： Wireless communication Training Uncertainty Large language models Symbols Artificial neural networks Linguistics Natural language processing Calibration Demodulation

来源：评论

学校读者我要写书评

暂无评论

Using magnetic tunnel junctions as unconventional computing devices

Using magnetic tunnel junctions as unconventional computing ...

引用

2023 IEEE International Magnetic Conference - Short Papers, INTERMAG Short Papers 2023

作者： Grimaldi, Andrea Mazza, Luciano Tullo, Pietro Rodrigues, Davi Crupi, Vincenza Carpentieri, Mario Puliafito, Vito Finocchio, Giovanni University of Messina Department of Mathematical and Computer Sciences Physical Sciences and Earth Sciences MessinaI-98166 Italy Polytechnic of Bari Department of Electrical and Information Engineering BariI-70125 Italy

ISBN: (纸本)9798350338362

Hardware for specialized computing tasks is becoming more and more relevant in computational research due to the nearing intrinsic physical limitations of Von Neumann architecture. Most of the industrially relevant problems belong to the combinatorial optimization class and require a nondeterministic approach to be solved efficiently. To this end, magnetic tunnel junctions (MTJs) are one of the best candidate devices as building blocks to implement unconventional computation paradigms thanks to their tunable stochasticity, their size, and their compatibility with current CMOS technologies. Here, we present an MTJ-based implementation of two state-of-the-art paradigms, evaluate them and compare them over relevant computational problems. © 2023 IEEE.

关键词： Combinatorial optimization

来源：评论

学校读者我要写书评

暂无评论

Contextual Boosting to Explainable SVM Classification 1

引用

Fuzzy Logic and Technology, and Aggregation Operators 13th Conference of the European Society for Fuzzy Logic and Technology, EUSFLAT 2023, and 12th International Summer School on Aggregation Operators, AGOP 2023

作者： Loor, Marcelo Tapia-Rosero, Ana De Tré, Guy Department of Telecommunications and Information Processing Ghent University Sint-Pietersnieuwstraat 41 Ghent9000 Belgium Department of Electrical and Computer Engineering ESPOL Polytechnic University Campus Gustavo Galindo Guayaquil Ecuador

ISBN: (数字)9783031399657

ISBN: (纸本)9783031399640

Finding suitable mechanisms whereby rationale behind support vector machine (SVM) predictions can be known and understood without substantial difficulties is an ongoing challenge. Aiming to find such a mechanism, we look into the contextualization of SVM models. Hence, we propose a novel explainable SVM classifier that makes use of a parallel arrangement of contextualized SVM models for offering predictions that depend on a particular event, situation or idea. The proposed classifier allows decision makers to state in a clear manner the context of the predictions they would like to be offered. This aspect is deemed to be important since decision makers can take advantage of the improvement in the interpretability of such contextualized predictions for making more informed decisions. The improvement in interpretability is illustrated through an example in which digitized handwritten vowels are contextually identified. Another example where hand gestures are recognized by means of electromyography (EMG) signals shows how the proposed classifier can also improve the accuracy of the resulting models. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

An Attention-based Self-Supervised Approach to Monocular Depth Estimation from UAV captured Video Sequences

An Attention-based Self-Supervised Approach to Monocular Dep...

引用

Pattern Recognition and Artificial Intelligence (PRAI), International Conference on

作者： Yasmina Benkhoui Tahar El-Korchi Reinhold Ludwig Electrical and Computer Engineering Department Worcester Polytechnic Institute Civil and Environmental Engineering Department Worcester Polytechnic Institute

UAV-based inspections represent a promising solution to contactless civil infrastructure condition assessment, especially bridges. In this context, depth estimation is crucial as it helps predict the distance to objects, plan flight trajectories, and avoid obstacles. Recently, learning-based methods made significant progress in monocular depth estimation. However, supervised approaches require a large amount of annotated depth data which remains challenging to collect. As an alternative, researchers have been investigating self-supervised learning, where the depth estimation network is trained by projecting a view to nearby views and minimizing the re-projection photometric loss. However, current methods lack accuracy and robustness since they do not exploit structural information, leading to inaccurate spatial representations and surface discontinuities. In addition, monocular depth estimation suffers from scale ambiguity, and external or prior information is needed for scale recovery. This work sought to overcome these issues by adapting the standard joint network of estimating depth and pose based on a self-learned depth prediction network from monocular videos. The proposed method introduces a soft attention mechanism and a recurrent block. The attention mechanism guides the neural network’s attention to the global structures and local attributes and recovers the spatial relationship of visual features, whereas the recurrent block leverages sequence frames to capture temporal information in addition to spatial information. Our experiments and ablation study on AeroBridge dataset demonstrate that these modifications efficiently capture spatio-temporal information and generate competitive qualitative and quantitative results compared to state-of-the-art methods confirming the practical applicability of our design in inspection tasks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-Robot Scan-n-Print for Wire Arc Additive Manufacturing

引用

ASME Letters in Translational Robotics 2025年第1期1卷 011003页

作者： Lu, Chen-Lung He, Honglu Ren, Jinhan Dhar, Joni Saunders, Glenn Julius, Agung Samuel, Johnson Wen, John T. https://***/01rtyzb94Rensselaer Polytechnic Institute Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute Troy NY 12180 Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute Troy NY 12180 https://***/01rtyzb94Rensselaer Polytechnic Institute Department of Mechanical Aerospace and Nuclear Engineering Rensselaer Polytechnic Institute Troy NY 12180 https://***/01rtyzb94Rensselaer Polytechnic Institute Manufacturing Innovation Center Rensselaer Polytechnic Institute Troy NY 12180 Department of Mechanical Aerospace and Nuclear Engineering Rensselaer Polytechnic Institute Troy NY 12180

Robotic Wire Arc Additive Manufacturing (WAAM) is a metal additive manufacturing technology offering flexible 3D printing while ensuring high-quality near-net-shape final parts. However, WAAM also suffers from geometric imprecision, especially for low-melting-point metals such as aluminum alloys. In this article, we present a multi-robot framework for WAAM process monitoring and control. We consider a three-robot setup: a 6-DoF welding robot, a 2-DoF trunnion platform, and a 6-DoF sensing robot with a wrist-mounted laser line scanner measuring the printed part height profile. The welding parameters, including the wire feed rate, are held constant based on the materials used, so the control input is the robot path speed. The measured output is the part height profile. The planning phase decomposes the target shape into slices of uniform height. During runtime, the sensing robot scans each printed layer, and the robot path speed for the next layer is adjusted based on the deviation from the desired profile. The adjustment is based on an identified model correlating the path speed to changes in height. The control architecture coordinates the synchronous motion and data acquisition between all robots and sensors. Using a three-robot WAAM testbed, we demonstrate significant improvements of the closed-loop scan-n-print approach over the current open loop result on both a flat wall and a more complex turbine blade shape.

关键词： welding process control multiple robot scan-n-print welding robot WAAM

来源：评论

学校读者我要写书评

暂无评论

Guest Editorial Enabling Technologies and Systems for Industry 5.0: From Foundation Models to Foundation Intelligence

引用

IEEE Transactions on Systems, Man, and Cybernetics: Systems 2024年第11期54卷 6496-6499页

作者： Tang, Ying Tian, Yonglin Lin, Yilun Lv, Chen Fanti, Maria Pia Rowan University Department of Electrical and Computer Engineering GlassboroNJ08028 United States Chinese Academy of Sciences State Key Laboratory of Multimodal Artificial Intelligence Systems Beijing100190 China Institute of Automation Chinese Academy of Sciences Beijing100190 China Shanghai AI Laboratory Urban Computing Laboratory Shanghai200232 China Nanyang Technological University School of Mechanical and Aerospace Engineering Jurong West639798 Singapore Polytechnic University of Bari Department of Electrical and Information Engineering Bari70126 Italy

来源：评论

学校读者我要写书评

暂无评论

Adaptive DDS-PLL Beamsteering Architecture based on Real-Time Angle-of-Arrival Estimation

Adaptive DDS-PLL Beamsteering Architecture based on Real-Tim...

引用

Midwest Symposium on Circuits and Systems (MWSCAS)

作者： Antonello Florio Giuseppe Coviello Claudio Talarico Gianfranco Avitabile Department of Electrical and Information Engineering Polytechnic University of Bari Bari BA Italy Department of Electrical and Computer Engineering Gonzaga University Spokane WA

ISBN: (数字)9798350387179

ISBN: (纸本)9798350387186

When dealing with millimeter wave communications, one way to mitigate the path loss contribution involves the use of beam steering and beamforming techniques, which allow to improve the gain of the antenna system. Particularly interesting is the case of adaptive beamformers, which are able to steer the antenna radiation pattern towards the position of a given target in real-time. In this work, we propose a novel architecture for adaptive beam steering based on real-time Angle-of-Arrival (AoA) estimation. The phase shift is introduced in the Local Oscillator (LO) path thanks to a modified DDS-PLL structure. The main equations are presented, along with a validation campaign involving Hardware-in-the-Loop (HiL) simulations.

关键词： Local oscillators Beam steering Power demand Hardware-in-the-loop simulation Millimeter wave technology Estimation Prototypes

来源：评论

学校读者我要写书评

暂无评论

Compact Two-Stage Origami Horn Antenna for Terahertz CubeSat Networks

Compact Two-Stage Origami Horn Antenna for Terahertz CubeSat...

引用

2022 IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, AP-S/URSI 2022

作者： Alqaraghuli, Ali J. Singh, Arjun Jornet, Josep M. Northeastern University Department of Electrical and Computer Engineering BostonMA United States SUNY Polytechnic Institute Department of Engineering UticaNY United States

ISBN: (数字)9781665496582

ISBN: (纸本)9781665496582

With the rapid proliferation of high-speed satellite communication networks enabled by dense constellations in space, CubeSats are expected to be more compact than ever while maintaining large bandwidth operation. A candidate enabler for high-data-rate CubeSat constellations is the terahertz (THz) band (0.1-10 THz), which supports both compactness and large bandwidth. These two come at the cost of high propagation losses and impose the use of very high gain directional antennas, which on their turn drastically impact otherwise simple tasks such as neighbor discovery and broadcasting with nearby satellites. To simultaneously alleviate the challenges of neighbor discovery while overcoming the high path losses in long links, a THz origami horn antenna utilizing commercially-available WR4 waveguide is presented. The proposed antenna has two stages supporting phases of neighbor discovery and high-data-rate communication. The two stages demonstrate gains of 15.7 dBi and 30.7 dBi, respectively, at 218 GHz. The horn antenna is explored as a solution for enabling CubeSat swarms at terahertz frequencies. © 2022 IEEE.

关键词： Horn antennas

来源：评论

学校读者我要写书评

暂无评论

Analysis and Treatment of Missing Data in Happiness Score Using Machine Learning 21

Analysis and Treatment of Missing Data in Happiness Score Us...

引用

21st IEEE International Conference on Smart Communities: Improving Quality of Life using AI, Robotics and IoT, HONET 2024

作者： Touati, Wassim El Touati, Yamen Hadj-Alouane, Nejib Ben Jemai, Abderrazek Polytechnic School of Tunisia Carthage University SERCOM Laboratory Tunis Tunisia Northern Border University Faculty of Computing and Information Technology Department of Computer Science Saudi Arabia ENIT University of Tunis EI Manar OASIS Laboratory Tunisia American University in Dubai Electrical Engineering and Computer Engineering Department Dubai United Arab Emirates

ISBN: (纸本)9798350378078

Machine learning, a vital part of artificial intelli-gence, improves our ability to make predictions from complex data. The success of these predictions relies heavily on the model's fit with its data and the data's consistency. Accurate predictions depend on closely matching the model to the data patterns. However, missing data can seriously disrupt the accuracy of predictions, causing errors. To address this, we recommend adopting a proposed framework that enhances predictions by effectively filling in missing data. Our objective is to refine the accuracy of our predictions, especially in scenarios where the data might be incomplete and features exhibit patterns of dependency. In the context of missing data, our approach concentrates on enhancing regression performance to minimize the effects of missing data as effectively as possible. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

G-Band Metamaterial-Based Circulator for FCC-Compliant Space-to-Earth Communication 32

G-Band Metamaterial-Based Circulator for FCC-Compliant Space...

引用

32nd International Symposium of Space Terahertz Technology, ISSTT 2022

作者： Alqaraghuli, Ali J. Aliaga, Sergi Singh, Arjun Jornet, Josep M. Department of Electrical and Computer Engineering Northeastern University BostonMA United States Department of Engineering SUNY Polytechnic Institute UticaNY United States

A metamaterial-based circulator operating at G-band is presented for the purpose of high-data-rate space-to-Earth communication. The circulator is designed to be equipped on a CubeSat using a single antenna, integrating a metamaterial within an interface of three WR3 waveguides measuring 3 mm in length per waveguide. The isolation is achieved between transmitting (downlink) signals at 295 GHz and reception (uplink) at 218 GHz. The frequencies are chosen in accordance with FCC regulations and aim to provide a high-data-rate download link for CubeSats deployed in non-terrestrial networks. © 2022 32nd International Symposium of Space Terahertz Technology, ISSTT 2022 All rights reserved.

关键词： Metamaterials

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：