检索结果-内蒙古大学图书馆

8th NAFOSTED Conference on Information and computer Science, NICS 2021

作者： Nguyen, Son Truong Tuyen, Nguyen Quang Phuc, Nguyen Hong Hanoi University of Science and Technology School of Mechanical Engineering Viet Nam University of Aizu School of Computer Science and Engineering Japan Eastern International University Department of Software Engineering Viet Nam

ISBN: (纸本)9781665410014

Recently, style transfer is a research area that attracts a lot of attention, which transfers the style of an image onto a content target. Extensive research on style transfer has aimed at speeding up processing or generating high-quality stylized images. Most approaches only produce an output from a content and style image pair, while a few others use complex architectures and can only produce a certain number of outputs. In this paper, we propose a simple method for representing style features in many ways called Deep Feature Rotation (DFR), while not only producing diverse outputs but also still achieving effective stylization compared to more complex methods. Our approach is representative of the many ways of augmentation for intermediate feature embedding without consuming too much computational expense. We also analyze our method by visualizing output in different rotation weights. Our code is available at https://***/sonnguyen129/deep-feature-rotation. © 2021 IEEE.

关键词： Rotation

来源：评论

学校读者我要写书评

暂无评论

Everything you wanted to know about ChatGPT: Components, capabilities, applications, and opportunities

引用

Internet technology Letters 2024年第6期7卷

作者： Heidari, Arash Navimipour, Nima Jafari Zeadally, Sherali Chamola, Vinay Department of Software Engineering Haliç University Istanbul Turkey Department of Computer Engineering Faculty of Engineering and Natural Science İstanbul Atlas University Istanbul Turkey Department of Computer Engineering Faculty of Engineering and Natural Sciences Kadir Has University Istanbul Turkey Future Technology Research Center National Yunlin University of Science and Technology Douliou Taiwan College of Communication and Information University of Kentucky LexingtonKY United States Pilani India

Conversational Artificial Intelligence (AI) and Natural Language Processing have advanced significantly with the creation of a Generative Pre-trained Transformer (ChatGPT) by OpenAI. ChatGPT uses deep learning techniques like transformer architecture and self-attention mechanisms to replicate human speech and provide coherent and appropriate replies to the situation. The model mainly depends on the patterns discovered in the training data, which might result in incorrect or illogical conclusions. In the context of open-domain chats, we investigate the components, capabilities constraints, and potential applications of ChatGPT along with future opportunities. We begin by describing the components of ChatGPT followed by a definition of chatbots. We present a new taxonomy to classify them. Our taxonomy includes rule-based chatbots, retrieval-based chatbots, generative chatbots, and hybrid chatbots. Next, we describe the capabilities and constraints of ChatGPT. Finally, we present potential applications of ChatGPT and future research opportunities. The results showed that ChatGPT, a transformer-based chatbot model, utilizes encoders to produce coherent responses. © 2024 John Wiley & Sons Ltd.

关键词： Taxonomies

来源：评论

学校读者我要写书评

暂无评论

Energy Storage and Forecasting Error Impact Analysis In Photovoltaic Equipped Residential Nano-Grids

Energy Storage and Forecasting Error Impact Analysis In Phot...

引用

Compatibility in Power Electronics (CPE)

作者： Hossein Nourollahi Hokmabad Oleksandr Husev Juri Belikov Dmitri Vinnikov Eduard Petlenkov Department of Electrical Power Engineering and Mechatronics Tallinn University of Technology Tallinn Estonia Department of Software Science Tallinn University of Technology Tallinn Estonia Department of Computer Systems Tallinn University of Technology Tallinn Estonia

Battery energy storage systems (BESS) enable many applications for photovoltaic (PV) equipped nano-grids. Stored excessive energy is utilized for energy arbitrage, demand response during blackouts, and peak shaving. This technology helped utility service providers deal with the duck-curve-effect and intermittency of renewable energy systems. In this paper, we investigate the benefits of using energy storage systems in PV nano-grids for residential sectors and evaluate the relationship between battery presence and system performance. We also analyze the effects of battery degradation on energy independence and self-sufficiency and investigate the relationship between simple payback time and state of health (SoH) levels of nano-grid components. Additionally, we propose a simple rule-based energy supervision strategy and evaluate its performance under various forecasting error levels, highlighting the importance of high-performance forecasting methods for proper energy management systems. Our contributions will help select and optimize PV-battery solutions and energy management algorithms to achieve the highest benefit and minimize investment risks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Adaptive Deep Reinforcement Learning-based Resource Management for Complex Decision Making in Industry Internet of Things Applications

引用

Procedia computer Science 2025年 252卷 231-240页

作者： Niharika Karne Chandra Shekar Ramagundam Ranga Rao Patnala Shashishekhar Ramagundam Sai Nitisha Tadiboina Sr. Principal Systems Engineer in C&RS Platform services DT Amtrak USA University of south alabama Department of electrical engineering Masters in electrical engineering Senior Software Engineer in Cloud platform services City Of Philadelphia Pennsylvania USA Vice president Neuberger Berman US Software Development & Engineering in Content Discovery & AI Comcast USA Troy University Montgomery Campus AL. Department of computer Science Masters In Computer Science Department of Information Technology GEICO - Chevy Chase Chevy Chase MD 20815 United States

An independent industrial system may transform into a connected network through the assistance of Industrial Internet of Things (IIoT). The deployed sensors in the IIoT maintain surveillance of the industrial machinery and equipment. As a result, safety and reliability emerge as the primary concerns in IIoT. This presents a variety of well-known and increasing issues related to the industrial system. The IIoT devices are exposed to a wide range of malware, threats, and assaults. To prevent the IIoT devices from malware effects, effective protection plans must be implemented. But adequate security mechanisms are not be incorporated in IIoT devices with limited resources. It is essential to ensure the accuracy and dependability of information gathered by IIoT devices. Decisions taken with incomplete or inaccurate data might be devastating. To overcome these difficulties deep learning with reinforcement learning for complex decision-making in industry applications is developed in this research work. In this developed model, an Adaptive Deep Reinforcement learning (ADRL)-based resource management is performed to reduce the operation cost associated with IIoT deployments. Energy efficiency is essential in IIoT ecosystem, particularly for the devices that run on batteries. Through dynamic resource allocation based on workload needs and energy limits, ADRL-based resource management optimizes the usage of energy. The reliability of the designed model is enhanced by fine-tuning the parameters from DRL using the Ship Rescue Optimization (SRO) algorithm. Thus, ADRL-based resource management systems make real-time decisions based on current environmental conditions and system requirements. This helps the IIoT systems to react quickly to change demands and optimize resource allocation. Finally, the experimental analysis is performed to find the success rate of the developed resource management system via various metrics. Throughout the validation, the statistical analysis of the

关键词： Resource Management Industrial Internet of Things Applications Complex Decision Making Adaptive Deep Reinforcement Learning Ship Rescue Optimization

来源：评论

学校读者我要写书评

暂无评论

Multi-Scale Dilated Convolution Network for Long-Term Time Series Forecasting

arXiv

引用

arXiv 2024年

作者： Li, Feifei Guo, Suhan Han, Feng Zhao, Jian Shen, Furao National Key Laboratory for Novel Software Technology Nanjing University China Department of Computer Science and Technology Nanjing University China School of Artificial Intelligence Nanjing University China School of Electronic Science and Engineering Nanjing University China

Accurate forecasting of long-term time series has important applications for decision making and planning. However, it remains challenging to capture the long-term dependencies in time series data. To better extract long-term dependencies, We propose Multi-Scale Dilated Convolution Network (MSDCN), a method that utilizes a shallow dilated convolution architecture to capture the period and trend characteristics of long time series. We design different convolution blocks with exponentially growing dilations and varying kernel sizes to sample time series data at different scales. Furthermore, we utilize traditional autoregressive model to capture the linear relationships within the data. To validate the effectiveness of the proposed approach, we conduct experiments on eight challenging long-term time series forecasting benchmark datasets. The experimental results show that our approach outperforms the prior state-of-the-art approaches and shows significant inference speed improvements compared to several strong baseline methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Patient-Level Anatomy Meets Scanning-Level Physics: Personalized Federated Low-Dose CT Denoising Empowered by Large Language Model

arXiv

引用

arXiv 2025年

作者： Yang, Ziyuan Chen, Yingyu Wang, Zhiwen Shan, Hongming Chen, Yang Zhang, Yi College of Computer Science Sichuan University China Key Laboratory of Data Protection and Intelligent Management Ministry of Education Sichuan University China School of Cyber Science and Engineering Sichuan University China School of Computer Science and Software Engineering Southwest Petroleum University China Institute of Science and Technology for Brain-inspired Intelligence Fudan University China School of Computer Science and Engineering Southeast University China

Reducing radiation doses benefits patients, however, the resultant low-dose computed tomography (LDCT) images often suffer from clinically unacceptable noise and artifacts. While deep learning (DL) shows promise in LDCT reconstruction, it requires large-scale data collection from multiple clients, raising privacy concerns. Federated learning (FL) has been introduced to address these privacy concerns;however, current methods are typically tailored to specific scanning protocols, which limits their generalizability and makes them less effective for unseen protocols. To address these issues, we propose SCAN-PhysFed, a novel SCanning- and ANatomy-level personalized Physics-Driven Federated learning paradigm for LDCT reconstruction. Since the noise distribution in LDCT data is closely tied to scanning protocols and anatomical structures being scanned, we design a dual-level physics-informed way to address these challenges. Specifically, we incorporate physical and anatomical prompts into our physics-informed hypernetworks to capture scanning- and anatomy-specific information, enabling dual-level physics-driven personalization of imaging features. These prompts are derived from the scanning protocol and the radiology report generated by a medical large language model (MLLM), respectively. Subsequently, client-specific decoders project these dual-level personalized imaging features back into the image domain. Besides, to tackle the challenge of unseen data, we introduce a novel protocol vector-quantization strategy Copyright © 2025, The Authors. All rights reserved.

关键词： computerized tomography

来源：评论

学校读者我要写书评

暂无评论

Enhancing Abstractive Dialogue Summarization with Internal Knowledge

Enhancing Abstractive Dialogue Summarization with Internal K...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Shaolei Wang Hanhan Ma Yunxiang Zhang Jing Ma Liang He School of Software Xinjiang University Urumqi China School of Computer Science and Technology Xinjiang University Urumqi China School of Intelligence Science and Technology Xinjiang University Urumqi China Department of Electronic Engineering Tsinghua University Beijing China

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

The task of dialogue summarization involves distilling a given dialogue into a concise and coherent summary. However, discrepancies in language styles between dialogues and summaries, scattered crucial information, incomplete utterances with ellipsis, and coreferences bring unique challenges to dialogue summarization. To tackle these challenges, we present multiple strategies in this study to extract crucial information with varying levels of granularity in dialogue from word-level and utterance-level semantics. This crucial information is used as valuable annotations on the dialogue text to train the model to recognize and utilize key information during the training phase, and enhance the model’s ability to identify crucial information during inference to generate better dialogue summaries. Experimental results on the SAMSum and DialogSum datasets shows that our method outperforms strong baseline models in terms of both ROUGE and BERTScore metrics. We corroborate these improvements through human evaluation.

关键词： Training Measurement Knowledge engineering Text recognition Annotations Semantics Neural networks

来源：评论

学校读者我要写书评

暂无评论

Bi-Orthogonal Projection Learning for Dimensionality Reduction

Bi-Orthogonal Projection Learning for Dimensionality Reducti...

引用

Intelligent Power and Systems (ICIPS), International Conference on

作者： Yingyi Liang Junjun Fan Chunsheng Li Jiajun Wen Guangdong University of Technology Guangzhou CN College of Computer Science & Software Engineering Shenzhen University Shenzhen CN Laboratory for Artificial Intelligence in Design Hong Kong

This paper proposes a novel bi-orthogonal projection learning (BOPL) for dimensionality reduction (DR) methods, which further extends the existing DR to a more flexible, robust, and sparse embedding framework. Unlike conventional methods that use only one kind of projection for learning in DR, the proposed BOPL introduces two kinds of projection, which are orthogonal and provides the true projection the freedom for more accurate data transformation in subspace learning. Inspired by the observation that the two projections share many similar data structures, the projections are expected to preserve the similarity structure of data by using two different reconstruction ways. The proposed method can also handle data corrupted by noises since a sparse item is employed to compensate for the noises during DR. Several novel unsupervised DR methods (i.e., BOPL_PCA, BOPL_NPE, and BOPL_LPP) are derived from the proposed framework. The results of experiments on the natural and synthetic data sets demonstrate that the proposed methods outperform the existing well-established DR methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The Compiler of DFC: A Source Code Converter that Transform the Dataflow Code to the Multi-threaded C Code 1

引用

21st International Conference on Parallel and Distributed Computing, Applications, and Technologies, PDCAT 2020

作者： Du, Zheng Zhang, Jing Li, Jinrong Du, Haixin Shu, Jiwu Luo, Qiuming Department of Computer Science and Technology Tsinghua University Beijing China NHPCC/Guangdong Key Laboratory of Popular HPC College of Computer Science and Software Engineering Shenzhen University Shenzhen China

ISBN: (数字)9783030692445

ISBN: (纸本)9783030692438

The working principle of DFC compiler is introduced in this article. DFC is a grammatical extension of standard C language, with special DF function which describe the dependence of computing DAG. DFC compiler, dfcc, is used to convert the DFC source codes to standard C codes with the assistance of multi-threaded library. The lexical rules and grammatical rules are used to setup the AST of DFC codes, and then it is converted to the AST of standard C without DF function nodes. The derived AST is printed into a text file, a normal C file, and finally is processed by GCC to obtain the executable file. The experiment gives some demonstration of that converting details, and the memory footprint is studied to show an ideal scalability. © 2021, Springer Nature Switzerland AG.

关键词： Program compilers

来源：评论

学校读者我要写书评

暂无评论

Enhancing Convergence in Federated Learning: A Contribution-Aware Asynchronous Approach

arXiv

引用

arXiv 2024年

作者： Xu, Changxin Qiao, Yuxin Zhou, Zhanxin Ni, Fanghao Xiong, Jize Computer Information Technology Northern Arizona University Flagstaff United States Big Data and Business Intelligence Universidad Internacional Isabel I de Castilla Burgos Spain Software Engineering Indian Institute of Technology Assam Guwahati India

Federated Learning (FL) is a distributed machine learning paradigm that allows clients to train models on their data while preserving their privacy. FL algorithms, such as Federated Averaging (FedAvg) and its variants, have been shown to converge well in many scenarios. However, these methods require clients to upload their local updates to the server in a synchronous manner, which can be slow and unreliable in realistic FL settings. To address this issue, researchers have developed asynchronous FL methods that allow clients to continue training on their local data using a stale global model. However, most of these methods simply aggregate all of the received updates without considering their relative contributions, which can slow down convergence. In this paper, we propose a contribution-aware asynchronous FL method that takes into account the staleness and statistical heterogeneity of the received updates. Our method dynamically adjusts the contribution of each update based on these factors, which can speed up convergence compared to existing methods. © 2024, CC BY.

关键词： Machine 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：