检索结果-内蒙古大学图书馆

Annual International Conference of the IEEE engineering in Medicine and Biology Society (EMBC)

作者： Quoc Bao Phan Linh Nguyen Ngoc Thang Bui Dinh C. Nguyen Lan Zhang Tuy Tan Nguyen School of Informatics Computing and Cyber Systems Northern Arizona University Flagstaff AZ USA Department of Radiology Mayo Clinic Rochester MN USA Department of Electrical and Computer Engineering The University of Alabama in Huntsville Huntsville AL USA

ISBN: (数字)9798350371499

ISBN: (纸本)9798350371505

This paper presents a novel approach for classifying electrocardiogram (ECG) signals in healthcare applications using federated learning and stacked convolutional neural networks (CNNs). Our innovative technique leverages the distributed nature of federated learning to collaboratively train a high-performance model while preserving data privacy on local devices. We propose a stacked CNN architecture tailored for ECG data, effectively extracting discriminative features across different temporal scales. The evaluation confirms the strength of our approach, culminating in a final model accuracy of 98.6% after 100 communication rounds, significantly exceeding baseline performance. This promising result paves the way for accurate and privacy-preserving ECG classification in diverse healthcare settings, potentially leading to improved diagnosis and patient monitoring.

关键词： Data privacy Accuracy Protocols Federated learning Scalability Medical services Electrocardiography Feature extraction Robustness Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous Treatment Effect Estimation with Subpopulation Identification for Personalized Medicine in Opioid Use Disorder

arXiv

引用

arXiv 2024年

作者： Lee, Seungyeon Liu, Ruoqi Song, Wenyu Zhang, Ping Department of Computer Science and Engineering The Ohio State University United States Department of Biomedical Informatics The Ohio State University United States Department of Medicine Brigham and Women’s Hospital Harvard Medical School United States

Deep learning models have demonstrated promising results in estimating treatment effects (TEE). However, most of them overlook the variations in treatment outcomes among subgroups with distinct characteristics. This limitation hinders their ability to provide accurate estimations and treatment recommendations for specific subgroups. In this study, we introduce a novel neural network-based framework, named SubgroupTE, which incorporates subgroup identification and treatment effect estimation. SubgroupTE identifies diverse subgroups and simultaneously estimates treatment effects for each subgroup, improving the treatment effect estimation by considering the heterogeneity of treatment responses. Comparative experiments on synthetic data show that SubgroupTE outperforms existing models in treatment effect estimation. Furthermore, experiments on a real-world dataset related to opioid use disorder (OUD) demonstrate the potential of our approach to enhance personalized treatment recommendations for OUD patients. Copyright © 2024, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

A Study of Multilayer Omega Networks Operating with Cut-Through Mechanism under Uniform Traffic 5

A Study of Multilayer Omega Networks Operating with Cut-Thro...

引用

5th South-East Europe Design Automation, computer engineering, computer Networks and Social Media Conference, SEEDA-CECNSM 2020

作者： Stergiou, Eleftherios Garofalakis, John Liarokapis, Dimitrios Margariti, Spiridoula V. University of Ioannina Dept. Informatics and Telecommunications Arta Greece Dept. of Computer Engineering and Informatics Patras Greece

ISBN: (纸本)9781728164458

There is an enduring interest in developing more efficient multistage interconnection networks (MINs), as they are a key aspect of various switching sectors and the demands for higher data transfer rates. Here, semi-layer Omega type networks operating under the cut-through forwarding policy are selected as a test-bed subject for performance evaluation. This special shuffle-exchange network was modelled, analysed, and simulated. Moreover, this type of fabric is compared with corresponding single-layer Omega networks operating via cut-through or a 'store and forward' forwarding technique. The results show, in a quantitative manner, how superior is the under study fabric compared to the corresponding single-layer constructions, in terms of network size, showing the advantages that are provided by both usages: fan-out and cut-through mechanisms. Moreover, the roles of the switch element's size and the reliability metric were investigated. The findings that are derived can be valuable assets for MIN designers, in order to achieve more efficient delivering networks with low overall deployment costs. © 2020 IEEE.

关键词： Data transfer

来源：评论

学校读者我要写书评

暂无评论

Implementation voice command system for soccer robot ersow

Implementation voice command system for soccer robot ersow

引用

2020 International Electronics Symposium, IES 2020

作者： Prasetyo, Muhammad Rizal Wibowo, Iwan Kumianto Bachtiar, Mochamad Mobed Priambudi, Renardi Adryantoro Anwar, Khoirul Segara, Putu Bagus Kertha Politeknik Elektronika Negeri Surabaya Department of Informatics and Computer Engineering Surabaya Indonesia

ISBN: (数字)9781728195308

ISBN: (纸本)9781728195308

ERSOW is a wheeled soccer robot that is included in the Middle Size League (MSL) category in the Indonesian Wheeled Robot Soccer Contest division (Wheeled KRSBI). Wheeled soccer robot has Artificial Intelligent (AI) for kick the ball, receive the ball, feed the ball, recognize the ball, recognize the opponent, recognize the goal, receive instructions from the base station, and so forth. This research focuses on giving instructions to ERSOW through the base station using the voice command system. The system uses speech as input in the form of analog signals. Speech recognition is done by using a deep speech package so as to produce output in the form of text. The system will run on the Robot Operating System (ROS). The result of speech recognition using 13 trained speakers when tested by one speaker in different distance show average Word Error Rate (WER) 0.46% and Word Accuracy (W Acc) is 99.54%. When tested by five different speaker using trained speakers show average WER is 4.37% and W Acc is 95.63%, when using non trained speakers show average WER is 28.25% and W Acc is 71.75%. Base station implementation shows the simulation of the robot when the user gives instruction by his or her own voice. © 2020 IEEE.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Instance-wise Grasp Synthesis for Robotic Grasping

Instance-wise Grasp Synthesis for Robotic Grasping

引用

IEEE International Conference on Robotics and Automation (ICRA)

作者： Yucheng Xu Mohammadreza Kasaei Hamidreza Kasaei Zhibin Li School of Informatics University of Edinburgh UK Department of Artificial Intelligence Faculty of Science and Engineering Bernoulli Institute University of Groningen The Netherlands Department of Computer Science University College London UK

Generating high-quality instance-wise grasp con-figurations provides critical information of how to grasp specific objects in a multi-object environment and is of high importance for robot manipulation tasks. This work proposed a novel Single-Stage Grasp (SSG) synthesis network, which performs high-quality instance-wise grasp synthesis in a single stage: instance mask and grasp configurations are generated for each object simultaneously. Our method outperforms state-of-the-art on robotic grasp prediction based on the OCID-Grasp dataset, and performs competitively on the JACQUARD dataset. The benchmarking results showed significant improvements compared to the baseline on the accuracy of generated grasp configurations. The performance of the proposed method has been validated through both extensive simulations and real robot experiments for three tasks including single object pick-and-place, grasp synthesis in cluttered environments and table cleaning task.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-Label Classification of Film Genres Based on Synopsis Using Support Vector Machine, Logistic Regression and Naïve Bayes Algorithms

Multi-Label Classification of Film Genres Based on Synopsis ...

引用

International Conference on Information Technology, Information Systems and Electrical engineering (ICITISEE)

作者： Jihadul Akbar Ema Utami Ainul Yaqin Magister of Informatics Engineering Universitas Amikom Yogyakarta Yogyakarta Indonesia Computer Science Faculty Universitas Amikom Yogyakarta Yogyakarta Indonesia

Movies were still a very popular means of entertainment. The current distribution of internet users causes a large amount of movie data to be created and distributed online. The emergence of movie streaming services makes consumers very interested in using automatic film genre classification. In this study, a multi-label film genre classification will be carried out based on an English synopsis. Data were collected from the Internet Movie Database (IMDb) website. The amount of data used in this study was 10,432 lines of data obtained using scraping techniques on June 7, 2022. Researchers divided the dataset labels into 18 labels representing each genre. Feature extraction using TF-IDF and Stemming. The multi-label classification algorithm used is the Support Vector Machine, Logistic Regression, and Naive Bayes Algorithms. Optimal parameter search using GridSearch of each algorithm. The optimum result in this study was obtained f1-score value of 0.58 using the SVM algorithm with TF-IDF feature extraction with stemming dataset, followed by NB with the f1-score value of 0.48 and LR with an f1-score value of 0.43.

关键词： Electrical engineering Support vector machine classification Entertainment industry Motion pictures Feature extraction Classification algorithms Naive Bayes methods

来源：评论

学校读者我要写书评

暂无评论

AI-based Models to Predict the Heart Rate Using PPG and Accelerometer Signals During Physical Exercise

AI-based Models to Predict the Heart Rate Using PPG and Acce...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Lucas Ribeiro Hélder P. Oliveira Xiao Hu Tania Pereira INESC TEC INESC TEC - Institute for Systems and Computer Engineering Technology and Science FEUP - Faculty of Engineering University of Porto Porto Portugal INESC TEC - Institute for Systems and Computer Engineering Technology and Science FCUP - Faculty of Science University of Porto Porto Portugal Department of Biomedical Informatics School of Medicine Department of Computer Science College of Arts and Sciences Nell Hodgson Woodruff School of Nursing Emory University Atlanta USA INESC TEC - Institute for Systems and Computer Engineering Technology and Science Porto Portugal

PPG signal is a valuable resource for continuous heart rate monitoring; however, this signal suffers from artifact movements, which is particularly relevant during physical exercise and makes this biomedical signal difficult to use for heart rate detection during those activities. The purpose of this study was to develop learning models to determine heart rate using data from wearables (PPG and acceleration signals) and dealing with noise during physical exercise. Learning models based on CNNs and LSTMs were developed to predict the heart rate. The PPG signal was combined with data from accelerometers trying to overcome the noise movement on the PPG signal. Two datasets were used on this work: the 2015 IEEE Signal Processing Cup (SPC) dataset was used for training and testing, and another dataset was used for validation of the learning model (PPG-DaLiA dataset). The predictions obtained by the learning model represented a mean average error of 7.033±5.376 bpm for the SCP dataset, while a mean average error of 9.520±8.443 bpm for the validation set. The use of acceleration data increases the performance of the learning models on the prediction of the heart rate, showing the benefits of using this source of data to overcome the noise movement problem on the PPG signal. The combination of PPG signal with acceleration data could allow the learning models to use more information regarding the motion artifacts that affect the PPG and improve performance on the physiological event detections, which will largely spread the use of wearables on the healthcare applications for continuous monitor the physiological state allowing early and accurate detection of pathological events.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Heading Calibration in Robot Soccer ERSOW using Line Landmark on the Field

Heading Calibration in Robot Soccer ERSOW using Line Landmar...

引用

2020 International Electronics Symposium, IES 2020

作者： Lui Hakim Ihsan, Fadl Adryantoro Priambudi, Renardi Mobed Bachtiar, Mochamad Kumianto Wibowo, Iwan Politeknik Elektronika Negeri Surabaya Department of Informatics and Computer Engineering Surabaya Indonesia

ISBN: (纸本)9781728195308

This paper will present the heading calibration which is used on the Ersow's robot using line sensors. The robot must have the capabilities of avoiding obstacles, kicks the ball, and localization. All of those abilities are dependent on the robot heading data in the form of an angular degree unit. The drawback of the heading data is getting worse when the sensor is activated in a long period. This happens because IMU sensor is drifting along the time that causes a shifting orientation perception. This condition will create an enormous error in the robot system. The proposed way to solve this problem is calibrating the IMU data each necessary. The robot is placed in the field by looking for unique features. In this case, is the horizontal lines. We are using a microcontroller-based line sensor with an independent embedded microcontroller. The purposes of using the line sensors are the fast computation and the cheapest hardware to be built. The line sensor is expected to produce deviation data that is eligible to send towards Artificial Intelligent processed. Based on the result of the experiment average error by drifting robot heading-1.67 degree in every minute. The average error value with robot speed at 10 cm/s is 2.1 degree, at 50 cm/s is 1 degree, at 100 cm/s is 0.7 degree, and at 150 cm/s is 1.2 degree. The accuracy of this calibration method obtained as 99.65%. This improvement will make the robot increasing the accuracy of position estimation in the field. © 2020 IEEE.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Dynamic Local Ball Tracking in Middle Size League Robot Soccer ERSOW based on Kaiman Filter

Dynamic Local Ball Tracking in Middle Size League Robot Socc...

引用

2020 International Electronics Symposium, IES 2020

作者： Bachtiar, Mochamad Mobed Wibowo, Iwan Kurnianto Dikarinata, Rangga Priambudi, Renardi Adryantoro Anwar, Khoirul Politeknik Elektronika Negeri Surabaya Department of Informatics and Computer Engineering Surabaya Indonesia

ISBN: (纸本)9781728195308

The Robot Soccer uses the vision system to look for the ball continuously. The quality of vision object detection is the main factor that considered by the robot. Beside the quality, the performance of the detection process also affects the robot performance. The object detection is the heaviest process in entire ERSOW's robot process. In this paper, we addressed the ways optimizing the vision object detection process that enhanced by the tracking method using Kaiman Filter. The Kaiman filter is also widely used for robotic purposes. The object has been equipped with a local ROI around them to limit the scanning on the entire frame when detection method is running. The local ROI will reduce the computation process and keeping the process in the sufficient resources that processor can handle. The Kaiman filter will predicted the object position and the direction of the object by considered the previous position and the times was taken. The Kaiman filter will lock the object and will follow the object without using detection feature anymore. From the results of tests conducted, the predicted value in several position has showed promising result. The average error on x-Axis is 1.425 pixels and in y-Axis 1.7226 pixels. This system can also reduce the average computation time from 31.67 Ms into 20.4 Ms. This research is expected to overcome the overwhelmed of the ERSOW's computation and increased the performance of the robot © 2020 IEEE.

关键词： Kalman filters

来源：评论

学校读者我要写书评

暂无评论

ActivityMamba: A CNN-Mamba hybrid neural network for efficient human activity recognition

引用

IEEE Transactions on Mobile Computing 2025年

作者： Luo, Fei Li, Anna Jiang, Bin Khan, Salabat Wu, Kaishun Wang, Lu Great Bay University School of Computing and Information Technology Dongguan China Lancaster University School of Computing & Communications United Kingdom University of Petroleum College of Oceanography and Space Informatics Qingdao China Shenzhen University College of Computer Science and Software Engineering Shenzhen China Guangzhou China

Current research in human activity recognition primarily emphasizes enhancing accuracy, with limited exploration into computational efficiency and hardware compatibility. Recently, Mamba has sparked substantial interest within the realm of deep learning. Mamba is a hardware-aware algorithm enabling very efficient training and inference. Researchers are applying Mamba to various tasks, demonstrating significant promise in both language and vision tasks. It is worthwhile to investigate the use of Mamba for efficient human activity recognition. In this paper, we proposed a hybrid neural network that integrates CNN and visual Mamba, called ActivityMamba. The SE-Mamba block in ActivityMamba utilizes both CNN's local and Mamba's global context modeling while keeping computation and memory efficiency. We evaluated the ActivityMamba on five public benchmark datasets collected by using three different sensing techniques. ActivityMamba achieved higher performance than vision transformers, vision Mamba, and CNNs with fewer FLOPs and parameters. It sets a new SOTA on all five datasets, which are 91.78% OA and 89.13% F1 on the USC-HAD dataset, 99.19% OA and 98.64% F1 on the UT-HAR dataset, 99.82% OA and F1 on the DIAT dataset, 98.59% OA and 98.65% F1 on the UCI-HAR dataset, and 95.41% OA and 93.14% F1 on the UniMib dataset. Our work is the first to investigate the CNN-Mamba hybrid network for efficient human activity recognition. © 2002-2012 IEEE.

关键词： Active 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：