检索结果-内蒙古大学图书馆

Deep Learning in Heart Sound Analysis:From Techniques to Clinical Applications

Health data science 2024年第1期4卷 88-109页

作者： Qinghao Zhao Shijia Geng Boya Wang Yutong Sun Wenchang Nie Baochen Bai Chao Yu Feng Zhang Gongzheng Tang Deyun Zhang Yuxi Zhou Jian Liu Shenda Hong Department of Cardiology Peking University People’s HospitalBeijingChina HeartVoice Medical Technology HefeiChina Key Laboratory of Carcinogenesis and Translational Research(Ministry of Education/Beijing) Department of Gastrointestinal OncologyPeking University Cancer Hospital and InstituteBeijingChina National Institute of Health Data Science Peking UniversityBeijingChina Institute of Medical Technology Health Science Center of Peking UniversityBeijingChina Department of Computer Science Tianjin University of TechnologyTianjinChina DCST BNRistRIITInstitute of Internet IndustryTsinghua UniversityBeijingChina

Importance:Heart sound auscultation is a routinely used physical examination in clinical practice to identify potential cardiac abnormalities. However, accurate interpretation of heart sounds requires specialized training and experience, which limits its generalizability. Deep learning, a subset of machine learning, involves training artiffcial neural networks to learn from large datasets and perform complex tasks with intricate patterns. Over the past decade, deep learning has been successfully applied to heart sound analysis, achieving remarkable results and accumulating substantial heart sound data for model training. Although several reviews have summarized deep learning algorithms for heart sound analysis, there is a lack of comprehensive summaries regarding the available heart sound data and the clinical applications. Highlights:This review will compile the commonly used heart sound datasets, introduce the fundamentals and state-of-the-art techniques in heart sound analysis and deep learning, and summarize the current applications of deep learning for heart sound analysis, along with their limitations and areas for future improvement. Conclusions:The integration of deep learning into heart sound analysis represents a signiffcant advancement in clinical practice. The growing availability of heart sound datasets and the continuous development of deep learning techniques contribute to the improvement and broader clinical adoption of these models. However, ongoing research is needed to address existing challenges and reffne these technologies for broader clinical use.

关键词： Deep specialized routine

来源：评论

学校读者我要写书评

暂无评论

AI-Enhanced Resource Allocation for LPWAN-based LoRaWAN:A Hybrid TinyML and Deep Learning Approach

引用

IEEE Internet of Things Journal 2025年

作者： Lodhi, Muhammad Ali Sun, Xiaobing Mahmood, Khalid Lodhi, Anum Park, Youngho Hussain, Majid Yangzhou University School of Information Engineering Yangzhou China National Yunlin University of Science and Technology Graduate School of Intelligent Data Science Douliu64002 Taiwan University of Electronic Science and Technology of China School of computer science and engineering China Kyungpook National University School of Electronics Engineering Daegu41566 Korea Republic of The University of Faisalabad Department of Computer Science Faisalabad38850 Pakistan

The integration of Artificial Intelligence (AI) with Low Power Wide Area Networks (LPWAN) offers a promising approach to address resource constraints and dynamic network conditions inherent in these networks. However, deploying complex AI algorithms on resource-limited edge devices presents significant challenges due to their limited computational capabilities. In this study, we propose a hybrid Tiny Machine Learning (TinyML) and Deep Neural Network (DNN)-based solution for optimizing resource allocation in LPWAN-based LoRaWAN networks, targeting both static and mobile applications. Our approach leverages the strengths of a 1-D Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) model implemented on the network server, combined with TinyML models deployed on edge devices. The CNN-LSTM model predicts optimal spreading factor and transmission power by analyzing spatial and temporal patterns from real-time data, while the TinyML models enable edge devices to autonomously adjust communication parameters in resource-constrained and disconnected scenarios. This hybrid framework enhances network performance by improving the packet success ratio (PSR), maximizing energy efficiency, and addressing the challenges posed by dynamic IoT environments. © 2014 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Enhancing Computational Model for Human Motion Recognition Using Deep Learning

Enhancing Computational Model for Human Motion Recognition U...

引用

Intelligent Signal Processing and Effective Communication Technologies (INSPECT), IEEE International Conference on

作者： Ram Kumar Yadav A. Daniel Birendra Kumar Apeksha Sakhare Rahul Moriwal Raju Pawar Department of Computer Science and Engineering Amity university Gwalior Gwalior India Department of Data Science Internet of Things and Cyber Security G H Raisoni college of engineering Nagpur India Department of Computer Science and Engineering G H Raisoni college of engineering Nagpur India

ISBN: (数字)9798350379525

ISBN: (纸本)9798350379532

Human motion recognition (HMR) is a fundamental task in computer vision with applications in healthcare, surveillance, human-computer interaction, and intelligent environments. This paper proposes a better-performing model for HMR using VGG16 and starts the preprocessing on the diverse dataset of RGB images. Our meticulous process involves normalization, resizing, and then extracting images. These preparatory steps lay the foundation for robust model training. Exploration then takes two VGG16, EfficientNetB7 and ensemble VGG16-EfficientNet B7 deep learning architectures.

关键词： Training Human computer interaction Deep learning computer vision Computational modeling Surveillance Medical services Signal processing Communications technology Human activity recognition

来源：评论

学校读者我要写书评

暂无评论

UFID: A Unified Framework for Black-box Input-level Backdoor Detection on Diffusion Models

arXiv

引用

arXiv 2024年

作者： Guan, Zihan Hu, Mengxuan Li, Sheng Vullikanti, Anil Kumar Department of Computer Science University of Virginia United States Biocomplexity Institute University of Virginia United States School of Data Science University of Virginia United States

Diffusion models are vulnerable to backdoor attacks, where malicious attackers inject backdoors by poisoning certain training samples during the training stage. This poses a significant threat to real-world applications in the Model-as-a-Service (MaaS) scenario, where users query diffusion models through APIs or directly download them from the internet. To mitigate the threat of backdoor attacks under MaaS, black-box input-level backdoor detection has drawn recent interest, where defenders aim to build a firewall that filters out backdoor samples in the inference stage, with access only to input queries and the generated results from diffusion models. Despite some preliminary explorations on the traditional classification tasks, these methods cannot be directly applied to the generative tasks due to two major challenges: (1) more diverse failures and (2) a multi-modality attack surface. In this paper, we propose a black-box input-level backdoor detection framework on diffusion models, called UFID. Our defense is motivated by an insightful causal analysis: Backdoor attacks serve as the confounder, introducing a spurious path from input to target images, which remains consistent even when we perturb the input samples with Gaussian noise. We further validate the intuition with theoretical analysis. Extensive experiments across different datasets on both conditional and unconditional diffusion models show that our method achieves superb performance on detection effectiveness and run-time efficiency. Copyright © 2024, The Authors. All rights reserved.

关键词： Black-box testing

来源：评论

学校读者我要写书评

暂无评论

Autonomous Multi-Objective Optimization Using Large Language Model

引用

IEEE Transactions on Evolutionary Computation 2025年

作者： Huang, Yuxiao Wu, Shenghao Zhang, Wenjie Wu, Jibin Feng, Liang Tan, Kay Chen Department of Data Science and Artificial Intelligence Hong Kong Polytechnic University Hong Kong Department of Electrical and Elctronics Engineering Hong Kong Polytechnic University Hong Kong Chongqing University College of Computer Science Chongqing Key Laboratory of Big Data Intelligence and Privacy Computing Chongqing400044 China

Multi-objective optimization problems (MOPs) are ubiquitous in real-world applications, presenting a complex challenge of balancing multiple conflicting objectives. Traditional multi-objective evolutionary algorithms (MOEAs), though effective, often rely on domain-specific expertise for improved optimization performance, hindering adaptability to unseen MOPs. In recent years, the Large Language Models (LLMs) has revolutionized software engineering by enabling the autonomous generation and refinement of programs. Leveraging this breakthrough, we propose a new LLM-based framework that autonomously designs MOEAs for solving MOPs. The proposed framework includes a robust testing module to refine the generated MOEA through error-driven dialogue with LLMs, a dynamic selection strategy along with informative prompting-based crossover and mutation to fit textual optimization pipeline. Our approach facilitates the design of MOEA without the extensive demands for expert intervention, thereby speeding up the innovation of MOEA. Empirical studies across various MOP categories validate the robustness and superior performance of our proposed framework. © 1997-2012 IEEE.

关键词： Multiobjective optimization

来源：评论

学校读者我要写书评

暂无评论

Wavelet Neuro-Fuzzy System (WNFS) in Predicting the Exchange Rate of the Rupiah Against the US Dollar

Wavelet Neuro-Fuzzy System (WNFS) in Predicting the Exchange...

引用

2021 International Conference on Artificial Intelligence and Big data Analytics, ICAIBDA 2021

作者： Muharam, Asep Budiyana Rambe, Muhamad Farid Ridho Sudanta, Muhammad Risqullah Hidayat, Akik Padjadjaran University Research Center for AI and Big Data Department of Computer Science Jatinangor Indonesia Padjadjaran University Department of Computer Science Jatinangor Indonesia

ISBN: (纸本)9781665408905

In In this digital era, online trading is increasingly in demand and is often the community's first choice for investment. One type of online trading that is often chosen is Foreign Exchange (Forex) which is a buying and selling transaction of foreign currency. Forex investment is a type of investment that could return (more returns but also has a high level of risk because the fluctuations in currency values are very dynamic and significant. Therefore, in this study, a prediction technique for the IDR/USD exchange rate will be built. to be able to make decisions in the investment step using a combination of models wavelet, neural networks, and fuzzy systems called the Neuro-Fuzzy Wavelet model. The process in the neuro-fuzzy wavelet model includes the decomposition of Discrete Wavelet Transform (DWT), Projection scale values resulting from discrete wavelet transformations and a significant number of lags become inputs for the neural network, perform backward elimination on the lag variable, cluster data using Fuzzy C-Means, and conduct learning using a neural network. Prediction results using Wavelet Neuro-Fuzzy in determining the prediction of the rupiah exchange rate against the dollar American shows an average accuracy of 95%. © 2021 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Visual-Language Alignment for Background Subtraction

Visual-Language Alignment for Background Subtraction

引用

IEEE International Conference on Multimedia and Expo Workshops (ICMEW)

作者： Jiahe Liu Dandan Zhu Sajid Javed China University of Petroleum (Beijing) China Beijing Key Laboratory of Petroleum Data Mining China Department of Computer Science Khalifa University of Science and Technology UAE

ISBN: (数字)9798350379815

ISBN: (纸本)9798350379822

Background Subtraction (BGS) is a fundamental task in video analysis, critical for many application scenarios. Despite the development of various methods to address the identification of moving objects, current techniques fall short when faced with the intricate challenges inherent in real-world settings. Two such challenges that persist are the presence of dynamic backgrounds, where the environmental backdrop is constantly changing, and camera jitter, which introduces erratic movements into the scene. In the field of computer vision, we introduce for the first time a vision-language model designed for BGS tasks, utilizing the integration of linguistic and visual information to enhance the understanding and interpretation of complex scenes within the context of background sub-traction efforts. Our model has been rigorously tested across three categories within the extensive CDNet-2014 dataset, the results indicate a compelling average F-measure of 0.9771, highlighting the model's proficiency. This investigation offers a new perspective and a novel solution for BGS, particularly in complex video scenarios.

关键词： Adaptation models Analytical models Visualization Computational modeling Jitter Cameras Transformers

来源：评论

学校读者我要写书评

暂无评论

Mirror Descent on Reproducing Kernel Banach Spaces

arXiv

引用

arXiv 2024年

作者： Kumar, Akash Belkin, Mikhail Pandit, Parthe Department of Computer Science and Engineering University of California San Diego United States Department of Computer Science and Engineering Halıcıoğlu Data Science Insititute University of California San Diego United States Indian Institute of Technology Bombay India

Recent advances in machine learning have led to increased interest in reproducing kernel Banach spaces (RKBS) as a more general framework that extends beyond reproducing kernel Hilbert spaces (RKHS). These works have resulted in the formulation of representer theorems under several regularized learning schemes. However, little is known about an optimization method that encompasses these results in this setting. This paper addresses a learning problem on Banach spaces endowed with a reproducing kernel, focusing on efficient optimization within RKBS. To tackle this challenge, we propose an algorithm based on mirror descent (MDA). Our approach involves an iterative method that employs gradient steps in the dual space of the Banach space using the reproducing kernel. We analyze the convergence properties of our algorithm under various assumptions and establish two types of results: first, we identify conditions under which a linear convergence rate is achievable, akin to optimization in the Euclidean setting, and provide a proof of the linear rate;second, we demonstrate a standard convergence rate in a constrained setting. Moreover, to instantiate this algorithm in practice, we introduce a novel family of RKBSs with p-norm (p ≠ 2), characterized by both an explicit dual map and a kernel. © 2024, CC BY.

关键词： Banach spaces

来源：评论

学校读者我要写书评

暂无评论

An Experimental Study of Passive UAV Tracking with Digital Arrays and Cellular Downlink Signals

arXiv

引用

arXiv 2024年

作者： Sun, Yifei Yu, Chao Luo, Yan Han, Tony Xiao Tan, Haisheng Wang, Rui Lau, Francis C.M. The Department of Electrical and Electronic Engineering Southern University of Science and Technology Shenzhen518055 China The Department of Computer Science School of Computing and Data Science The University of Hong Kong Hong Kong Huawei Technologies Company Ltd. Shenzhen518063 China School of Artificial Intelligence and Data Science University of Science and Technology of China Hefei230027 China

Given the prospects of the low-altitude economy (LAE) and the popularity of unmanned aerial vehicles (UAVs), there are increasing demands on monitoring flying objects at low altitude in wide urban areas. In this work, the widely deployed long-term evolution (LTE) base station (BS) is exploited to illuminate UAVs in bistatic trajectory tracking. Specifically, a passive sensing receiver with two digital antenna arrays is proposed and developed to capture both the line-of-sight (LoS) signal and the scattered signal off a target UAV. From their cross ambiguity function, the bistatic range, Doppler shift and angle-of-arrival (AoA) of the target UAV can be detected in a sequence of time slots. In order to address missed detections and false alarms of passive sensing, a multi-target tracking framework is adopted to track the trajectory of the target UAV. It is demonstrated by experiments that the proposed UAV tracking system can achieve a meter-level accuracy. Copyright © 2024, The Authors. All rights reserved.

关键词： Long Term Evolution (LTE)

来源：评论

学校读者我要写书评

暂无评论

Large Model for Small data: Foundation Model for Cross-Modal RF Human Activity Recognition

arXiv

引用

arXiv 2024年

作者： Weng, Yuxuan Wu, Guoquan Zheng, Tianyue Yang, Yanbing Luo, Jun Department of Computer Science and Engineering Southern University of Science and Technology China Smart City Center Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology China College of Computer Science Sichuan University China College of Computing and Data Science Nanyang Technological University Singapore

Radio-Frequency (RF)-based Human Activity Recognition (HAR) rises as a promising solution for applications unamenable to techniques requiring computer visions. However, the scarcity of labeled RF data due to their non-interpretable nature poses a significant obstacle. Thanks to the recent breakthrough of foundation models (FMs), extracting deep semantic insights from unlabeled visual data become viable, yet these vision-based FMs fall short when applied to small RF datasets. To bridge this gap, we introduce FM-Fi, an innovative cross-modal framework engineered to translate the knowledge of vision-based FMs for enhancing RF-based HAR systems. FM-Fi involves a novel cross-modal contrastive knowledge distillation mechanism, enabling an RF encoder to inherit the interpretative power of FMs for achieving zero-shot learning. It also employs the intrinsic capabilities of FM and RF to remove extraneous features for better alignment between the two modalities. The framework is further refined through metric-based few-shot learning techniques, aiming to boost the performance for predefined HAR tasks. Comprehensive evaluations evidently indicate that FM-Fi rivals the effectiveness of vision-based methodologies, and the evaluation results provide empirical validation of FM-Fi’s generalizability across various environments. © 2024, CC BY.

关键词： Zero-shot 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：