检索结果-内蒙古大学图书馆

International Conference on Information Networking

作者： Jin Zhu Amiya Nayak School of Electrical Engineering and Computer Science University of Ottawa Canada

ISBN: (数字)9798331506940

ISBN: (纸本)9798331506957

The paper presents a study on task offloading migration decisions. The aim is to minimize the total latency of the offloaded computational tasks. Actor-Critic algorithm was employed to solve the problem, and real world taxi traces were used to train and test the model. The results indicate that the scheme of only migrating to neighboring servers performs well compared with the scheme of only migrating to any server. The model has good average rewards and converges faster, and has a degree of generalization ability.

关键词： Training Multi-access edge computing Computational modeling Deep reinforcement learning Servers Public transportation Convergence

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Survey of Floor Plan Image Analysis and Related Applications

引用

International Journal on Document Analysis and Recognition 2025年 1-19页

作者： Khade, Rasika Jariwala, Krupa Chattopadhyay, Chiranjoy Department of Computer Science and Engineering Sardar Vallabhbhai National Institute of Technology Surat Surat India School of Computing and Data Sciences FLAME University Pune India

This paper presents an in-depth exploration of architectural floor plan analysis through computational statistics, addressing the layout optimization and performance enhancement of buildings. Floor plans, which include elements like windows, walls, doors, and appliances, are pivotal for collaboration among architects, engineers, and clients, allowing for tailored designs based on client preferences. The study acknowledges the complexity and diversity of floor plans, which pose significant challenges for both human and machine recognition. Our research provides a thorough review of the evolution and current state of floor plan analysis, as well as a projection into its future. We systematically classify existing models into nine categories-dimensional, room-based, structural, retrieval, object-detection, graph, map, residential, and area-based-and assess them against criteria such as application domain, dataset, publication year, and reported performance metrics. The paper also discusses the practical implications of computational floor plan analysis, such as space optimization, energy efficiency, accessibility, safety, and comfort in the built environment. By processing and recognizing floor plan images, we can gain valuable insights and propose solutions to these issues. Our comprehensive review not only charts the progress in the field but also highlights the unresolved challenges and potential directions for future research. This includes the need for advanced models that can effectively process the intricate details of floor plans to further the goals of sustainable and functional architectural design. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Document Image Floor plan Floor plan analysis Floor plan image retrieval Image Retrieval Neural networks

来源：评论

学校读者我要写书评

暂无评论

Wikipedia Empowered Natural Language Interface for Web Search 25th

Wikipedia Empowered Natural Language Interface for Web Sear...

引用

25th International Conference on Web Information Systems engineering, WISE 2024

作者： Wang, Haofen Chowdhury, S. M. Mazharul Hoque Jin, Wei College of Design and Innovation Tongji University Shanghai China Department of Computer Science and Engineering University of North Texas Denton United States

ISBN: (纸本)9789819605781

Understanding queries in information retrieval (IR) is crucial. Tasks like query classification or clustering exist but may lack precision. Detailed goal descriptions, like human annotations, are vital for improving query understanding and evaluating document relevance. This paper addresses two problems: (1) how to simplify natural language questions so keyword-based search engines can understand them, and (2) how to make this process lightweight but effective. The first problem is solved by using a context graph to translate questions into keyword queries. The second uses Wikipedia for entity recognition, disambiguation, and relevance feedback. Experiments on 892 TREC questions show a 21.45% overall MRR increase, with 36.41% for "why" questions, compared to Bing, Google, and Yahoo. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Structured Query Language

来源：评论

学校读者我要写书评

暂无评论

GSArch: Breaking Memory Barriers in 3D Gaussian Splatting Training via Architectural Support 31

GSArch: Breaking Memory Barriers in 3D Gaussian Splatting Tr...

引用

31st IEEE International Symposium on High Performance computer Architecture, HPCA 2025

作者： He, Houshu Li, Gang Liu, Fangxin Jiang, Li Liang, Xiaoyao Song, Zhuoran Shanghai Jiao Tong University Department of Computer Science and Engineering Shanghai China Institute of Automation Chinese Academy of Sciences Beijing China

ISBN: (纸本)9798331506476

3D Gaussian Splatting (3DGS) introduces a novel methodology for representing scenes with anisotropic 3D Gaussian primitives, achieving exceptional quality and rendering speed in neural scene representation (NSR). However, the insufficient training speed of 3DGS limits its applicability in tasks that require online learning to perceive dynamic environments, such as autonomous driving and embodied intelligence. Although recent work, GSCore, has introduced a specialized accelerator for the rendering process of 3DGS, it overlooks the time-consuming backward propagation during 3DGS *** this paper, we propose GSArch, a hardware architecture designed to overcome memory barriers and boost the efficiency of 3DGS training. Through a thorough characterization of 3DGS training, we identify three root causes of inefficiency: redundant data loading from off-chip memory, time-consuming atomic write operations, and severe bank conflicts during on-chip buffer reading. To address these challenges, GSArch introduces three architectural innovations. First, acknowledging that Gaussians vary in shape and often span multiple pixels, with larger Gaussians causing more repetitive data loading, GSArch employs hybrid memory management. This approach categorizes Gaussians into 'hot' and 'cold' ones, storing hot Gaussians in a fast but small on-chip buffer to reduce redundant loading while minimizing hardware costs. Second, GSArch leverages the informativeness variability of Gaussians' gradients to filter out low-contribution gradients, significantly reducing atomic operations. Lastly, a rearrangement unit is designed to pack conflicting memory read requests into non-conflicting bundles. Our evaluation results demonstrate that GSArch achieves up to 6.49 × and 15.42 × speedups compared to Nvidia A100 and Jetson AGX Xavier, respectively, with substantially lower energy consumption and negligible image quality loss. © 2025 IEEE.

关键词： Gaussian distribution

来源：评论

学校读者我要写书评

暂无评论

External-feedback-coupled microring lasers on a III-V-on-Si photonic chip for reservoir computing

External-feedback-coupled microring lasers on a III-V-on-Si ...

引用

Smart Photonic and Optoelectronic Integrated Circuits 2025

作者： Niu, Yue Poon, Andrew W. Photonic Device Laboratory Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology Clear Water Bay Hong Kong

ISBN: (纸本)9781510684881

Photonic reservoir computing using a single neuron node exploits transient states at different times of a nonlinear oscillator. A semiconductor laser diode (LD) with injection carriers and emission photons oscillating in a laser cavity can work as a photonic reservoir. An LD coupled with an external feedback cavity increases the memory capacity and the nonlinearity of a reservoir. A microring laser cavity is equipped with additional nonlinearities with two counterpropagating lasing modes in a compact size. In this paper, we present our progress in the design and fabrication of an external-feedbackcoupled microring laser on a heterogeneously integrated III-V-on-Si platform to construct a photonic reservoir computing system. © 2025 SPIE.

关键词： Photonic integrated circuits

来源：评论

学校读者我要写书评

暂无评论

Codesign for Broadcast Addressing Biochip Towards Tamper-Resistance and Enhanced Reliability 38

Codesign for Broadcast Addressing Biochip Towards Tamper-Res...

引用

38th International Conference on VLSI Design, VLSID 2025

作者： Majumdar, Ritwika Datta, Piyali Chakraborty, Arpan Pal, Rajat Kumar Institute of Engineering & Management University of Engineering and Management Kolkata India University of Calcutta Department of Computer Science and Engineering Kolkata India Inside S&R Kolkata India

ISBN: (纸本)9798331522445

Digital Microfluidic Biochips (DMFB), famous for their low manufacturing cost and high responsiveness, have shown vulnerability to actuation sequence tampering attacks with the motive of manipulating assay results. Broadcast addressing is a specialized pin-mapping scheme whose highly programmable nature can be leveraged to expose the effect of stealthy actuation tampering attacks via networks. State-of-the-art research producing tamper-resistant outcome do not ensure feasible wire-routing. Moreover, higher electrode switching activity degrades reliability of a biochip. There must be a codesign considering all these design, security, and reliability issues together for attaining safe and potential deliverables. As per our knowledge, this paper introduces the first pin-mapping method for broadcast addressing biochips addressing the problems of tamper-resistance, wire-routing, and switching toggles. Experimental results show that our method improves the design significantly in terms of well-known tamper-resistance metrics, provides a pin-mapping solution with feasible wire-routing, and potentially reduces switching toggles which in turn enhance reliability of the chip. © 2025 IEEE.

关键词： Biochips

来源：评论

学校读者我要写书评

暂无评论

Quantum Inspired DC Optimal Power Flow Under Frequency Constraint

Quantum Inspired DC Optimal Power Flow Under Frequency Const...

引用

IEEE Texas Power and Energy Conference (TPEC)

作者： Madan Rana Magar Nga Nguyen Electrical Engineering and Computer Science University of Wyoming Laramie WY

ISBN: (数字)9798331541125

ISBN: (纸本)9798331541132

This study uses quantum-inspired techniques to ad-dress the DC optimal power flow problem considering frequency constraints. Although numerous analytical and data-driven meth-ods have been developed to solve DC-OPF under stability security constraints, this problem has never been addressed in the context of quantum computing. This paper proposed a novel algorithm based on quantum computing principles that improves both the accuracy and efficiency of solving DC-OPF solutions while dealing with the complexity of frequency stability requirements. This method simplifies the frequency security constraint by converting it into an inertia constraint, incorporating all variables from the frequency deviation equation. This transformed constraint is then integrated into the DC-OPF model. This method will be tested on the IEEE 14 bus system using the IBM qiskit simulation tool. The alternate direction method of multipliers is used as a quantum optimization technique and validates the results against classical methods. Our findings show that this quantum-inspired approach effectively optimizes power flow and maintains frequency stability, indicating its significant potential to enhance modern power systems for more resilient and efficient energy management.

关键词： Quantum computing Power system stability Frequency conversion Stability analysis Mathematical models Planning Security Energy management Optimization Load flow

来源：评论

学校读者我要写书评

暂无评论

An Open Source Implementation of Wi-Fi 7 Multi-Link Operation in OMNeT++

An Open Source Implementation of Wi-Fi 7 Multi-Link Operatio...

引用

Annual Conference on Wireless On Demand Network Systems and Services (WONS)

作者： Doğanalp Ergenç Falko Dressler School of Electrical Engineering and Computer Science TU Berlin Germany

ISBN: (数字)9783903176713

ISBN: (纸本)9798331522025

IEEE 802.11be Extremely High Throughput (EHT) improves the previous Wi-Fi generations with increased connectivity, throughput, and reliability. One of its key features, multi-link operation (MLO), enables using multiple links operating on the available 2.4, 5, and 6 GHz frequency bands simultaneously for better spectrum efficiency. This also comes with a flexible orchestration of these links to address the demands of evolving wireless networks, such as low latency and high reliability. Although several studies have already analyzed the benefits of MLO with custom implementations, researchers still lack an open source platform to develop their own MLO modes and deployment settings. In this paper, we introduce an open source MLO implementation in the popular simulation toolkit OMNeT++. We also present two examples of MLO modes, namely link aggregation and redundancy, to demonstrate their effectiveness.

关键词： Wireless networks Redundancy Fault tolerant systems Link aggregation Quality of service Throughput Low latency communication Wireless fidelity Network systems

来源：评论

学校读者我要写书评

暂无评论

NeuronQuant: Accurate and Efficient Post-Training Quantization for Spiking Neural Networks 25

NeuronQuant: Accurate and Efficient Post-Training Quantizati...

引用

30th Asia and South Pacific Design Automation Conference, ASP-DAC 2025

作者： Li, Haomin Liu, Fangxin Sun, Zewen Wang, Zongwu Huang, Shiyuan Yang, Ning Jiang, Li Department of Computer Science and Engineering Shanghai Jiao Tong University China Shanghai Qi Zhi Institute China Tianjin University China

ISBN: (纸本)9798400706356

Spiking neural networks (SNNs) are an alternative computational paradigm to artificial neural networks (ANNs) that have attracted attention due to their event-driven execution mechanisms, enabling extremely low energy consumption. However, a significant challenge and opportunity in SNNs is to optimize memory and compute costs while maintaining accuracy, thereby further reducing energy consumption. Model quantization has been proposed as a promising technique to improve the running efficiency via the number of data bits reduction. Whereas, this technique has yet to be well studied in the neuromorphic computing domain. The underlying reason is that the behaviors of SNNs are quite different from those of ANNs, making 1) the accuracy of SNNs usually sensitive to data precision, and 2) the introduction of a temporal dimension to characterize neuronal dynamics. In this paper, we present NeuronQuant, an accurate and energy-efficient quantization framework to reduce the precision of neurons while maintaining accuracy. The key insight is to design a post-training quantization method guided by the activity of neurons, efficiently reducing the bit-width of parameters based on local relationships within neurons. Additionally, a budget-aware mixed bit-width allocation strategy for the total model size enables the adaptive growth and narrowing of precision in each layer, leading to a mixed-precision quantization scheme of the desired size. Extensive evaluations demonstrate that NeuronQuant can achieve a compressed SNN with 5.2 bits on average and 1.51× power consumption reduction with a superior model accuracy which is quite impressive for SNN. Code is released at https://***/shieldforever/NeuronQuant. © 2025 Copyright is held by the owner/author(s). Publication rights licensed to ACM.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Long Short-Term Memory (LSTM) Networks For Stock Market Prediction

Long Short-Term Memory (LSTM) Networks For Stock Market Pred...

引用

2025 International Conference on Pervasive Computational Technologies, ICPCT 2025

作者： Gupta, Anish Tyagi, Lalit Kumar Singh Sisodia, Vibhash Chandigarh Group of Colleges Chandigarh Engineering College Department of Computer Science & Engineering Mohali India Sri Bhawani Niketan Institute of Technology and Management Jaipur India

ISBN: (纸本)9798331508685

Long Short-Term Memory (LSTM) networks have garnered interest for their ability to remember temporal relations in sequential input. This research employs LSTM networks to estimate prices, which is crucial for stock market predictions. First, we explain the basic concepts of LSTM networks, then we demonstrate how they may be used for time series forecasting by analyzing and interpreting sequential data. Thus, the study begins with a brief overview of LSTM architecture and its potential advantages over traditional models. They show how important feature engineering and data preparation are for stock market prediction using LSTMs. We also consider optimizing LSTM modeling assessment methodologies and hyper parameters. LSTM networks are used in single-step and multi-step stock market forecasting, according to this publication. We employ fundamental research on news emotions and economic indexes to examine how environmental influences affect model efficiency. We also suggest ways to reduce overfitting and data scarcity, which are LSTM's drawbacks for stock market prediction. © 2025 IEEE.

关键词： Marketplaces

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：