检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Wu, Jingqian Zhu, Shuo Wang, Chutian Shi, Boxin Lam, Edmund Y. Department of Electrical and Electronic Engineering The University of Hong Kong Pokfulam Hong Kong National Key Laboratory for Multimedia Information Processing and National Engineering Research Center of Visual Technology School of Computer Science Peking University Beijing100871 China

—Recent advancements in 3D Gaussian Splatting (3D-GS) have demonstrated the potential of using 3D Gaussian primitives for high-speed, high-fidelity, and cost-efficient novel view synthesis from continuously calibrated input views. However, conventional methods require high-frame-rate dense and high-quality sharp images, which are time-consuming and inefficient to capture, especially in dynamic environments. Event cameras, with their high temporal resolution and ability to capture asynchronous brightness changes, offer a promising alternative for more reliable scene reconstruction without motion blur. In this paper, we propose SweepEvGS, a novel hardware-integrated method that leverages event cameras for robust and accurate novel view synthesis across various imaging settings from a single sweep. SweepEvGS utilizes the initial static frame with dense event streams captured during a single camera sweep to effectively reconstruct detailed scene views. We also introduce different real-world hardware imaging systems for real-world data collection and evaluation for future research. We validate the robustness and efficiency of SweepEvGS through experiments in three different imaging settings: synthetic objects, real-world macro-level, and real-world micro-level view synthesis. Our results demonstrate that SweepEvGS surpasses existing methods in visual rendering quality, rendering speed, and computational efficiency, highlighting its potential for dynamic practical applications. Copyright © 2024, The Authors. All rights reserved.

关键词： Rendering (computer graphics)

来源：评论

学校读者我要写书评

暂无评论

Multi-Objective Optimization of Antenna Design Using an Improved Artificial Hummingbird Algorithm Based on Bernoulli Chaotic Sampling

Multi-Objective Optimization of Antenna Design Using an Impr...

引用

IEEE MTT-S International Microwave Workshop Series on Advanced Materials and Processes for RF and THz Applications (IMWS-AMP)

作者： Jin-Peng He Peng-Fei Qin Guan-Long Huang Zi-Yu Pang Kam-Weng Tam School of Electronics and Information Engineering Foshan University Foshan Guangdong China Guangdong Provincial Key Laboratory of Industrial Intelligent Inspection Technology Foshan University Foshan China Department of Electrical and Computer Engineering University of Macau Macau SAR China

ISBN: (数字)9798331520717

ISBN: (纸本)9798331520724

This paper presents an improved artificial hummingbird algorithm Based on Bernoulli chaotic map sampling (BCMS-IAHA) for the multi-objective optimization design of series-fed patch antenna. In proposed algorithm, the Bernoulli chaotic map sampling method is used to improve the uniform distribution of the initial population. By optimizing the foraging strategy of the artificial hummingbird algorithm, it is more suitable for the multi-objective optimization design of the antenna. The proposed BCMS-IAHA method is verified through multi-objective optimization of series-fed patch antenna. The results show that the proposed BCMS-IAHA method has better optimization ability than other algorithm such as differential evolution for the multi-objective optimization design of series-fed patch antenna.

关键词： Radio frequency Microwave antennas Patch antennas Design methodology Conferences Sampling methods Microwave theory and techniques Optimization

来源：评论

学校读者我要写书评

暂无评论

Taming "data-hungry" reinforcement learning? stability in continuous state-action spaces 24

Taming "data-hungry" reinforcement learning? stability in co...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Yaqi Duan Martin J. Wainwright Department of Technology Operations and Statistics Stern School of Business New York University New York NY Laboratory for Information and Decision Systems Statistics and Data Science Center and Department of Electrical Engineering and Computer Science and Department of Mathematics Massachusetts Institute of Technology Cambridge MA

ISBN: (纸本)9798331314385

We introduce a novel framework for analyzing reinforcement learning (RL) in continuous state-action spaces, and use it to prove fast rates of convergence in both off-line and on-line settings. Our analysis highlights two key stability properties, relating to how changes in value functions and/or policies affect the Bellman operator and occupation measures. We argue that these properties are satisfied in many continuous state-action Markov decision processes. Our analysis also offers fresh perspectives on the roles of pessimism and optimism in off-line and on-line RL.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A VISION FOR CLEANER RIVERS: HARNESSING SNAPSHOT HYPERSPECTRAL IMAGING TO DETECT MACRO-PLASTIC LITTER

arXiv

引用

arXiv 2023年

作者： Hanson, Nathaniel Demirkaya, Ahmet Erdoğmuş, Deniz Stubbins, Aron Padır, Taşkın Imbiriba, Tales Institute for Experiential Robotics Northeastern University BostonMA United States Department of Electrical and Computer Engineering Northeastern University BostonMA United States Marine Science Center Northeastern University BostonMA United States

Plastic waste entering the riverine harms local ecosystems leading to negative ecological and economic impacts. Large parcels of plastic waste are transported from inland to oceans leading to a global scale problem of floating debris fields. In this context, efficient and automatized monitoring of mismanaged plastic waste is paramount. To address this problem, we analyze the feasibility of macro-plastic litter detection using computational imaging approaches in river-like scenarios. We enable near-real-time tracking of partially submerged plastics by using snapshot Visible-Shortwave Infrared hyperspectral imaging. Our experiments indicate that imaging strategies associated with machine learning classification approaches can lead to high detection accuracy even in challenging scenarios, especially when leveraging hyperspectral data and nonlinear classifiers. All code, data, and models are available online: https://***/RIVeR-Lab/hyperspectr al macro plastic detection © 2023, CC BY.

关键词： Hyperspectral imaging

来源：评论

学校读者我要写书评

暂无评论

RigLSTM: Recurrent Independent Grid LSTM for Generalizable Sequence Learning

arXiv

引用

arXiv 2023年

作者： Wang, Ziyu Jiang, Wenhao Zhang, Zixuan Tang, Wei Yan, Junchi Department of Electrical and Computer Engineering National University of Singapore Singapore Guangming Laboratory Shenzheng China Department of Computer Science University of Illinois Urbana-Champaign United States The Department of Statistics University of British Columbia United Kingdom The Department of Computer Science and Engineering MoE Key Lab of Artificial Intelligence Shanghai Jiao Tong University Shanghai China

Sequential processes in real-world often carry a combination of simple subsystems that interact with each other in certain forms. Learning such a modular structure can often improve the robustness against environmental changes. In this paper, we propose recurrent independent Grid LSTM (RigLSTM), composed of a group of independent LSTM cells that cooperate with each other, for exploiting the underlying modular structure of the target task. Our model adopts cell selection, input feature selection, hidden state selection, and soft state updating to achieve a better generalization ability on the basis of the recent Grid LSTM for the tasks where some factors differ between training and evaluation. Specifically, at each time step, only a fraction of cells are activated, and the activated cells select relevant inputs and cells to communicate with. At the end of one time step, the hidden states of the activated cells are updated by considering the relevance between the inputs and the hidden states from the last and current time steps. Extensive experiments on diversified sequential modeling tasks are conducted to show the superior generalization ability when there exist changes in the testing environment. Source code is available at https://***/ziyuwwang/rig-lstm. Copyright © 2023, The Authors. All rights reserved.

关键词： Cells

来源：评论

学校读者我要写书评

暂无评论

Development of a compact mouse retinal imaging system using spectral domain optical coherence tomography (SD-OCT) with a full field stimulation channel for recording fast optoretinograms (ORGs) 34

Development of a compact mouse retinal imaging system using ...

引用

Ophthalmic Technologies XXXIV 2024

作者： Jo, Hang Chan Pijewska, Ewelina Meleppat, Ratheesh K. Jonnal, Ravi S. Kim, Dae Yu Zawadzki, Robert J. Department of Electrical and Computer Engineering Inha University 100 Inha-ro Michuhol-gu Incheon22212 Korea Republic of Dept. of Ophthalmology and Vision Science UC Davis Eye Center SacramentoCA95817 United States UC Davis Eye-Pod Small Animal Ocular Imaging Laboratory Department of Cell Biology and Human Anatomy University of California Davis 4320 Tupper Hall DavisCA95616 United States

ISBN: (纸本)9781510669079

While the a-wave of mouse electroretinogram (ERG) occurs within 50 milliseconds after exposure to light, the optoretinogram (ORG) slower than a 20Hz sampling rate could face limitations in observing immediate morphological changes from the retina. In this study, we develop a compact custom-built mouse ORG system based on spectral domain optical coherence tomography (SD-OCT) for 100Hz~1KHz B-scan rates comprised of 100 kHz A-scans. All the optics of the developed ORG system are designed on a 24 x 24 inches optical breadboard to move easily as well as to combine with the ERG system in a dark room. Without using a fundus camera, the OCT system provides en-face images from high-pass filtering and square of the OCT spectral signal for mouse retinal positioning in-vivo before acquiring ORG data. The 490nm LED for light stimulus is generated to make uniform illumination at the mouse retina using the Maxwellian view method. The common path of the OCT scanning light and the visible LED is built with achromatic doublet lens combinations based on optical simulation with Opticstudio® . The developed compact ORG system can not only observe light-evoked responses with 1~10 milliseconds but also be used for the studies of correlations between ORG and ERG in the mouse retina. © 2024 SPIE.

关键词： Optical tomography

来源：评论

学校读者我要写书评

暂无评论

Extreme Supervised Algorithm for Day Ahead Market Price Forecasting

Extreme Supervised Algorithm for Day Ahead Market Price Fore...

引用

IEEE International Smart Cities Conference (ISC2)

作者： Stylianos Loizidis Spyros Theocharides Venizelos Venizelou Demetris Evagorou Giorgos Makrides Andreas Kyprianou George E. Georghiou Department of Electrical and Computer Engineering PV Technology Laboratory FOSS Research Centre for Sustainable Energy University of Cyprus Nicosia Cyprus Energy Regulatory Authority Nicosia Cyprus

Deregulation of electricity markets has ushered in a new era of heightened competition, allowing for the inclusion of fresh market entrants. However, market participation bears challenges related the extremely high volatility of prices that are affected by too many interrelated factors, such as weather conditions, production, consumption, renewable production, fuel prices, unexpected socio-political and health events, etc. Therefore, the electricity market shows unexpected changes that can lead to very high (extremely high) or very low levels (negative) prices, exposing the participants to high financial risks. Given the stochastic nature of electricity energy prices, this work employs the Extreme Learning Machine in conjunction with the Bootstrap method for forecasting electricity prices for the next day. Two distinct cases are examined. In the first case, the prediction model is trained only with historical market price data, while in the second one with forecast data. The reason this is done is to see which data gives better forecasts. The methodology was applied to German and Finnish market data, 2019–2022.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Green Object Tracker (GOT) without Offline Pre-training

arXiv

引用

arXiv 2023年

作者： Zhou, Zhiruo You, Suya Kuo, C.-C. Jay The Ming-Hsieh Department of Electrical and Computer Engineering University of Southern California CA90089-2564 United States Army Research Laboratory AdelphiMD United States

Supervised trackers trained on labeled data dominate the single object tracking field for superior tracking accuracy. The labeling cost and the huge computational complexity hinder their applications on edge devices. Unsupervised learning methods have also been investigated to reduce the labeling cost but their complexity remains high. Aiming at lightweight high-performance tracking, feasibility without offline pre-training, and algorithmic transparency, we propose a new single object tracking method, called the green object tracker (GOT), in this work. GOT conducts an ensemble of three prediction branches for robust box tracking: 1) a global object-based correlator to predict the object location roughly, 2) a local patch-based correlator to build temporal correlations of small spatial units, and 3) a superpixel-based segmentator to exploit the spatial information of the target frame. GOT offers competitive tracking accuracy with state-of-the-art unsupervised trackers, which demand heavy offline pre-training, at a lower computation cost. GOT has a tiny model size ( Copyright © 2023, The Authors. All rights reserved.

关键词： Parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

Active IRS Aided Secure Design for Integrated Sensing and SWIPT

Active IRS Aided Secure Design for Integrated Sensing and SW...

引用

International Symposium on Microwave, Antenna, Propagation and EMC Technologies for Wireless Communications (MAPE)

作者： Jinlei Xu Jifa Zhang Mingqian Liu Nan Zhao Naofal Al-Dhahir Xianbin Wang School of Information and Communication Engineering Dalian University of Technology Dalian China State Key Laboratory of Integrated Service Networks Xidian University Xian Shanxi China Department of Electrical and Computer Engineering University of Texas at Dallas Richardson TX USA

ISBN: (数字)9798331533922

ISBN: (纸本)9798331533939

In this paper, an active intelligent reflecting surface (IRS) assisted secure integrated sensing and simultaneous wireless information and power transfer system is proposed with the power splitting (PS) model adopted. Aiming to maximize the harvested power while satisfying the constraints of the integrated sidelobe level ratio and secrecy rate, an optimization problem is formulated to jointly determine the appropriate transmit beamforming, artificial noise (AN) vectors, PS ratios, as well as the amplification factors and phase shifts of the active IRS, which is non-convex and intractable. To overcome this challenge, we leverage semidefinite programming to transform it into a convex counterpart, and propose an alternating optimization algorithm to obtain a solution for the original problem. Simulation results demonstrate the effectiveness of the proposed scheme compared to benchmarks.

关键词： Wireless communication Microwave antennas Wireless sensor networks Array signal processing Simulation Transforms Benchmark testing Sensors Simultaneous wireless information and power transfer Optimization

来源：评论

学校读者我要写书评

暂无评论

Enhancing Control of a 21-Level Modular Multilevel Converter through Grid-Based Model Predictive Current Control

Enhancing Control of a 21-Level Modular Multilevel Converter...

引用

2023 IEEE Asia Meeting on Environment and electrical engineering, EEE-AM 2023

作者： Tran Hung, Cuong Nguyen Nhat, Tung Pham Viet, Phuong Nguyen Xuan, Phuc Nguyen Quang, Thuan Nhat Hai, Nguyen Thuyloi University Faculty of Electrical & Electronics Engineering Ha Noi Viet Nam Ha Noi University of Science and Technology Department of Industrial Automation Ha Noi Viet Nam Key Laboratory for High Voltage Techniques Institute of Energy Ha Noi Viet Nam Hanoi University of Science and Technology Computer Science Department School of Information and Communications Technology Hanoi Viet Nam

ISBN: (纸本)9798350381061

The Modular Multilevel Converter (MMC) has a flexible structure, extending the number of voltage levels on the AC side. This structure has many advantages such as low switching frequency, and a smaller number of transistor valves. Therefore, the MMC converter makes it easy to apply control algorithms and is widely applied to high-voltage power grids. This paper presents the control method of an MMC converter on the AC side, with 10 Sub-modules (SM) in each phase to generate 21 voltage levels. The process is performed by the current control method based on Finite Control Set - Model Predictive Control (FCS-MPC). With the advantage of fast response, optimizing the converter's working process by predicting the working states of the converter, according to the model, the FCS-MPC control algorithm has been used in this model. The results of the proposed method are proven by using Matlab/Simulink software. The results have shown the requirements of the DC/AC converter such as sinusoidal current, and voltage with 21 voltages. © 2023 IEEE.

关键词： Model predictive control

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：