检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Khandate, Gagan Wang, Boxuan Park, Sarah Ni, Weizhe Palacios, Joaquin Lampo, Kathryn Wu, Philippe Ho, Rosh Chang, Eric Ciocarlie, Matei Dept. of Computer Science Columbia University New YorkNY10027 United States Dept. of Mechanical Engineering Columbia University New YorkNY10027 United States

Pre-training on large datasets of robot demonstrations is a powerful technique for learning diverse manipulation skills but is often limited by the high cost and complexity of collecting robot-centric data, especially for tasks requiring tactile feedback. This work addresses these challenges by introducing a novel method for pre-training with multi-modal human demonstrations. Our approach jointly learns inverse and forward dynamics to extract latent state representations, towards learning manipulation specific representations. This enables efficient fine-tuning with only a small number of robot demonstrations, significantly improving data efficiency. Furthermore, our method allows for the use of multi-modal data, such as combination of vision and touch for manipulation. By leveraging latent dynamics modeling and tactile sensing, this approach paves the way for scalable robot manipulation learning based on human demonstrations. Copyright © 2025, The Authors. All rights reserved.

关键词： Human robot interaction

来源：评论

学校读者我要写书评

暂无评论

Design and Performance Analysis of Grid-Connected Hybrid Renewable Systems

Design and Performance Analysis of Grid-Connected Hybrid Ren...

引用

IEEE International WIE Conference on Electrical and computer engineering (WIECON-ECE)

作者： Md. Atikur Rahman Reyad Arafat Ibne Ikrarm Md. Mobasher Karim Sheik Erfan Ahmed Himu Mehedi Hasan Shiam Tasnia Fahrin Choity Sk. Md. Golam Mostafa Department of Electrical and Electronic Engineering International Islamic University Chittagong Chattogram Bangladesh Department of Electrical and Mechanical Engineering Nagoya Institute of Technology Nagoya Japan Department of Computer Science and Engineering International Islamic University Chittagong Chattogram Bangladesh

Renewable energy development aims to meet the world's energy needs while replacing fossil fuels and moving to more renewable energy sources (RES). However, several different RES are tied together, taking into account the local load, and connected to add more energy to the current grid. Sadly, it requires a considerable amount of land. Yet the effects of climate change and the expanding global population both put food security in danger and the contest for scarce land resources has increased. In this regard, it has been claimed that agro-photovoltaic systems, which combine photovoltaics with plant growth, present a chance for the synergistic integration of renewable energy and food production. In this study, a methodology is shown where a grid-connected APV system is combined with other sustainable generation such as wind turbines (WT) and biomass generators. The Perturbation and Observation (P&O) based MPPT controller is implemented to harness the maximum power from APV. An energy management system is employed to use the stored energy in the battery and store the renewable energy generated by APV. To simulate the performance and effectiveness of grid -connected - RES-energy storage, a thorough study has been conducted. WT and BM connected to the existing power grid can generate an avg. of 3508 kW/day. and the APV system can produce avg of 50 kWh/day. The simulation results demonstrated that the suggested strategy works since the controller may use the most power feasible in both steady-state and diverse weather conditions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

High Definition P2P Streaming: Quantitative Schema Analysis

High Definition P2P Streaming: Quantitative Schema Analysis

引用

Advances in Computing, Communication and Applied Informatics (ACCAI), International Conference on

作者： Lakshmana Phaneendra Maguluri J. Raja S Basheer Ahamed Vinay Kumar Seeniappan Kaliappan D Chitra Devi Department of Computer Science and Engineering Koneru Lakshmaiah Education Foundation Vaddeswaram Andhra Pradesh India Department of Computer Science and Engineering Sri Manakula Vinyagar Engineering College Autonomous Institution Pondicherry University Puducherry Tamil Nadu India Department of Mathematics PSNA College of Engineering and Technology (Autonomous) Dindigul Tamil Nadu India Department of Electronics and Communication Engineering Graphic Era Deemed to be University Dehradun Uttarakhand Department of Mechanical Engineering Velammal Institute of Technology Chennai Tamil Nadu India Department of Computer Science and Engineering S.A. Engineering College Thiruverkadu Tamil Nadu India

Because of the proliferation of video streaming services, an increasing number of individuals have recently joined distribution networks. Streaming video solutions that follow the conventional concept of relying on the server and clients use a significant amount of the server’s available bandwidth. Many Peer-to-Peer (P2P) streaming techniques have been emerged in recent years in response to the growing demand for live and on-demand video streaming services delivered via wireless networks at reduced costs to the hosting providers. P2P computing is an innovative method for developing programmes that run over-dispersed networks. Traditional error correction techniques continue to be poorly adapted for video transmission, despite the growth of potentially problematic channels for video transmission, such as wireless and IP. The combination of these two reasons served as the fundamental influence for developing a new set of approaches to solve video transmission errors. These methods are known as error concealment. In this study, we present a multiple description coding technique that we call Flexible Spatial-Temporal (FST). It is more error-resistant than previous approaches, particularly when it comes to the loss of frames due to unrelated channels, which is one of the most common types of error that can occur. It offers an approach that blends geographical concealment with temporal concealment at the receiver so that the missing frames may be hidden more effectively. According to the results of the trials, the solution that was recommended is capable of achieving a decent level of watching performance of video over a P2P based wireless network.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Explainable Boosting Ensemble Methods for Intrusion Detection in Internet of Medical Things (IoMT) Applications

Explainable Boosting Ensemble Methods for Intrusion Detectio...

引用

Digital Futures and Transformative Technologies (ICoDT2), International Conference on

作者： Fatima Sohail Muhammad Asim Mukhtar Bhatti Muhammad Awais Aamna Iqtidar School of Electrical Engineering and Computer Science National University of Sciences and Technology Islamabad Pakistan Military College of Signals National University of Sciences and Technology Islamabad Pakistan College of Electrical and Mechanical Engineering National University of Sciences and Technology Islamabad Pakistan

ISBN: (数字)9798350391862

ISBN: (纸本)9798350391879

This research focuses on developing an Intrusion Detection System (IDS) for Internet of Medical Things (IoMT) applications, utilizing the CICIoMT-2024 dataset. It approaches the problem as a classification challenge, classifying traffic as either one of eighteen types of attacks or benign. The study analyzes the effectiveness of three boosting algorithms—XGBoost, AdaBoost, and CatBoost—in determining their accuracy for identifying various types of attack and benign traffic. The experimental results reveal that the XGBoost classifier achieves an accuracy of 95.01%, AdaBoost records an accuracy of 92.89%, and CatBoost attains an accuracy of 88.37%. Additionally, eXplainable Artificial Intelligence (XAI) is employed to enhance the interpretability of the machine learning models.

关键词： Accuracy Explainable AI Intrusion detection Internet of Medical Things Boosting Classification algorithms Ensemble learning

来源：评论

学校读者我要写书评

暂无评论

Solid High-Speed Synchronous Reluctance Rotor Enabled by Multi-Material Additive Manufacturing

Solid High-Speed Synchronous Reluctance Rotor Enabled by Mul...

引用

IEEE Energy Conversion Congress and Exposition, ECCE

作者： Dante Newman Patrick Faue FNU Nishanth Behzad Rankouhi Frank E. Pfefferkorn Dan J. Thoma Eric Severson Department of Electrical and Computer Engineering University of Wisconsin-Madison Madison WI USA Department of Mechanical Engineering University of Wisconsin-Madison Madison WI USA Department of Materials Science and Engineering University of Wisconsin-Madison Madison WI USA

Synchronous reluctance (SynR) motor technology is promising to realize rare-earth material-free electric machines. However, structural challenges limit operation speed and subsequently power density compared to machines with rare-earth permanent magnets. This paper proposes and investigates multi-material additive manufacturing (MMAM) as a key-enabler to realize power-dense SynR machines. It does so by guiding magnetic flux through a solid rotor component by selective placement of magnetic and non-magnetic materials to enable high-speed operation. To validate this concept, samples are manufactured using a MMAM process and experimentally characterized to assess the structural and magnetic properties that can be expected for the proposed rotors. The data is then used in a multi-physics modeling framework to explore the design space of new MMAM rotor concepts. The simulated results in this paper reveal that MMAM technology can enable a 4x increase in rotor speed, resulting in 400 % power density improvements. The MMAM rotors achieved tip speeds of approximately 300 m/s and rotational speeds over 55 kRPM at comparable efficiencies to conventional designs, despite the presence of existing MMAM geometry restrictions. This study ultimately demonstrates that MMAM technology has the potential to enhance SynR machine operation speed and power density, making it a valuable option for high-performance applications.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Noise-Injected Spiking Graph Convolution for Energy-Efficient 3D Point Cloud Denoising

arXiv

引用

arXiv 2025年

作者： Li, Zikuan Wu, Qiaoyun Zhang, Jialin Zhang, Kaijun Wang, Jun School of Computer Science and Technology Nanjing University of Aeronautics and Astronautics China College of Mechanical and Electrical Engineering Nanjing University of Aeronautics and Astronautics China School of Artificial Intelligence Anhui University China

Spiking neural networks (SNNs), inspired by the spiking computation paradigm of the biological neural systems, have exhibited superior energy efficiency in 2D classification tasks over traditional artificial neural networks (ANNs). However, the regression potential of SNNs has not been well explored, especially in 3D point cloud processing. In this paper, we propose noise-injected spiking graph convolutional networks to leverage the full regression potential of SNNs in 3D point cloud denoising. Specifically, we first emulate the noise-injected neuronal dynamics to build noise-injected spiking neurons. On this basis, we design noise-injected spiking graph convolution for promoting disturbance-aware spiking representation learning on 3D points. Starting from the spiking graph convolution, we build two SNN-based denoising networks. One is a purely spiking graph convolutional network, which achieves low accuracy loss compared with some ANN-based alternatives, while resulting in significantly reduced energy consumption on two benchmark datasets, PU-Net and PC-Net. The other is a hybrid architecture that combines ANN-based learning with a high performance-efficiency trade-off in just a few time steps. Our work lights up SNN’s potential for 3D point cloud denoising, injecting new perspectives of exploring the deployment on neuromorphic chips while paving the way for developing energy-efficient 3D data acquisition devices. Copyright © 2025, The Authors. All rights reserved.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

FSLNet: Leveraging Skeleton Cue Guided FBIC for Bionic Intelligent Surveillance Robot via Transformer

FSLNet: Leveraging Skeleton Cue Guided FBIC for Bionic Intel...

引用

IEEE International Conference on Robotics and Biomimetics

作者： Hai Liu Qiang Chen Zhibing Liu Yongjian Deng Zhaoli Zhang You-Fu Li National Engineering Research Center for E-Learning Central China Normal University Wuhan China College of Computer Science Beijing University of Technology Beijing China Department of Mechanical Engineering City University of Hong Kong Kowloon Hong Kong

ISBN: (数字)9798331509644

ISBN: (纸本)9798331509651

Fine-grained bird image classification (FBIC) plays a critical role in robotic vision tracking and monitoring of endangered bird species. Traditional surveillance robots often face major challenges when operating in complex environments, especially under extreme conditions involving occlusions and complex backgrounds. To overcome these limitations, we developed a novel bionic intelligent surveillance robot that incorporates FSLNet - a fine-grained skeletal learning network that integrates skeletal tokens into a Vision Transformer (ViT) framework. Unlike traditional CNN-based models, FSLNet leverages skeletal tokens to capture key invariant features, improving the ability to distinguish between visually similar bird species under challenging circumstances. The architecture of our model includes visual token construction, skeleton token construction, and a transformer-based backbone for thorough feature learning. Extensive experiments on CUB-200-2011 and NABirds datasets demonstrate that our method achieves superior performance compared to state-of-the-art approaches.

关键词： Visualization Accuracy Biological system modeling Surveillance Birds Transformers Feature extraction Skeleton Robots Image classification

来源：评论

学校读者我要写书评

暂无评论

Enhancing Soil Carbon Content Classification Using Transfer Learning and FrFT on Millimeter Wave Radar Array Signals with Varied Backgrounds

Enhancing Soil Carbon Content Classification Using Transfer ...

引用

International Conference on Industrial Artificial Intelligence (IAI)

作者： Di An Patrick Zhang YangQuan Chen Electrical Engineering & Computer Science University of California Merced Merced U.S.A. MESA Lab University of California Merced Merced U.S.A. Department of Mechanical Engineering University of California Merced Merced U.S.A.

ISBN: (数字)9798350356618

ISBN: (纸本)9798350356625

The precise quantification of soil carbon content is crucial for understanding soil health and its impact on climate change. This study investigates the classification of soil carbon content using millimeter wave radar array signals, focusing on a comparative analysis of Support Vector Machines (SVM), Continuous Wavelet Transform (CWT), and Fractional Fourier Transform (FrFT) methods. Furthermore, raw time and CWT-processed signals using SVM are used as baseline to establish the fundamental classification performance. The key contribution of this research is the innovative use of FrFT, where the transformed signals are converted into images and subsequently fed into GoogLeNet through transfer learning. This approach leverages the powerful feature extraction capabilities of GoogleNet to enhance classification accuracy. Experiments were conducted on a dataset comprising millimeter wave signals reflecting varied backgrounds with the primary focus on one specific distance measurement. Results provided show that high resolution millimeter wave signals achieved up to 100% validation accuracy at multiple fractional orders. The improved accuracy highlights the potential of integrating advanced signal transformation methods with deep learning models for soil carbon content classification.

关键词： Support vector machines Training Accuracy Continuous wavelet transforms Millimeter wave measurements Transfer learning Soil Millimeter wave radar Wavelet analysis Carbon

来源：评论

学校读者我要写书评

暂无评论

Scalable trapping of single nanosized extracellular vesicles using plasmonics

arXiv

引用

arXiv 2023年

作者： Hong, Chuchuan Ndukaife, Justus C. Department of Electrical and Computer Engineering Vanderbilt University NashvilleTN United States Interdisciplinary Materials Science and Engineering Vanderbilt University NashvilleTN United States Department of Mechanical Engineering Vanderbilt University NashvilleTN United States

Heterogeneous nanoscale particles released by cells known as extracellular vesicles (EVs) are actively investigated for early disease detection1, monitoring2, and advanced therapeutics3. Due to their extremely small size, the stable trapping of nano-sized EVs using diffraction-limited optical tweezers4 has been met with challenges. Plasmon-enhanced optical trapping can confine light to the nanoscale to generate tight trapping potentials. Unfortunately, a long-standing challenge is that plasmonic tweezers have limited throughput and cannot provide rapid delivery and trapping of particles at plasmonic hotspots while precluding the intrinsic plasmon-induced photothermal heating effect at the same time. We report our original geometry-induced electrohydrodynamic tweezers (GET) that generate multiple electrohydrodynamic potentials for the parallelized transport and trapping of single EVs in parallel within seconds while enhancing the imaging of single trapped EVs. We show that the integration of nanoscale plasmonic cavities at the center of each GET trap results in the parallel placement of single EVs near plasmonic cavities enabling instantaneous plasmon-enhanced optical trapping upon laser illumination without any detrimental heating effect for the first time. These non-invasive scalable hybrid nanotweezers open new horizons for high-throughput tether-free plasmon-enhanced single EV trapping and spectroscopy. Other potential areas of impact include nanoplastics characterization, and scalable hybrid integration for quantum photonics. Copyright © 2023, The Authors. All rights reserved.

关键词： Electrohydrodynamics

来源：评论

学校读者我要写书评

暂无评论

A Distributed Time-Varying Optimization Algorithm For Networked Lagrangian Agents Generating Continuous Control Torques

A Distributed Time-Varying Optimization Algorithm For Networ...

引用

American Control Conference (ACC)

作者： Yong Ding Hanlei Wang Jie Mei Wei Ren Department of Electrical and Computer Engineering University of California Riverside CA USA Science and Technology on Space Intelligent Control Laboratory Beijing Institute of Control Engineering Beijing China School of Mechanical Engineering Harbin Institute of Technology Shenzhen China

In this paper, the distributed time-varying optimization problem is investigated for networked Lagrangian systems with parametric uncertainties. Due to the usage of the signum function in the control torque design, there might exist chattering while implementing the distributed time-varying optimization algorithms for networked Lagrangian agents in the existing works. To this end, we design a distributed optimization algorithm that is capable of generating continuous control torques and achieving exact optimum tracking. A simulation is presented to validate the effectiveness of the proposed algorithm.

关键词：

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：