检索结果-内蒙古大学图书馆

IEEE inst.umentation and Measurement technology Conference

作者： Herti Miawarni Tri Arief Sardjono Eko Setijadi Wijayanti Mauridhi Hery Purnomo Dept. of Electrical Engineering Faculty of Intelligent Electrical and Informatics Technology (F-Electics) Institut Teknologi Sepuluh Nopember Surabaya Indonesia Dept. of Electrical Engineering Dept. of Biomedical Engineering Faculty of Intelligent Electrical and Informatics Technology (F-Electics) Institut Teknologi Sepuluh Nopember Surabaya Indonesia Dept. of Architecture Faculty of Engineering Universitas Diponegoro Semarang Indonesia Dept. of Electrical Engineering Dept. of Computer Engineering Faculty of Intelligent Electrical and Informatics Technology (F-Electics) Institut Teknologi Sepuluh Nopember University Center of Excellence on Artificial Intelligence for Healthcare and Society (UCE AIHeS) Surabaya Indonesia

Recently, many countries have been dealing with an increasing amount of elderly who are living alone. Unfortunately, fall incidents in the elderly tend to happen, and without proper dependable care would lead to fatal injuries. The elderly must be closely monitored under constant observation. Accidental falls are a leading cause of injury, mortality, and mobility limitations in the elderly. The expenses of national healthcare systems are also significantly impacted by accidental falls. These findings highlight the importance of investing heavily in the study and advancement of fall detection and intervention technologies. It is essential to recognize them early and provide fast assistance by developing technologies that can simultaneously improve the standard and the safety of the elderly's living environment. In this paper, we present a computer vision algorithm for falling incident detection using a low-cost camera and deep learning model. Ultimately, this algorithm can be used for the cases of elderly, in which falling risk is high and fatal. Furthermore, one self-developed dataset was used to validate the suggested approach experimentally. We used an OpenPose-based feature using a low-cost camera input and then classified each event or activity as either fall or non-fall. Classification tasks are carried out by Long Short-Term Memory (LSTM) with multiple validation metrics. The LSTM in this work is also optimized using Bayesian method. Finally, our approach to recognize falls has accuracy of 99.5%, which is distinct from other past works.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Generating Automated Assistance Mechanism in Android Programming Self-learning System Using Automatic Testing Tools

Generating Automated Assistance Mechanism in Android Program...

引用

Innovation and Intelligence for informatics, Computing, and Technologies (3ICT), International Conference o

作者： Yan Watequlis Syaifudin Dewi Yanti Liliana Andi Baso Kaswar Triana Fatmawati Nobuo Funabiki Indrazno Siradjuddin Dept. of Information Technology State Polytechnic of Malang Malang Indonesia Dept. of Informatics Engineering State Polytechnic of Jakarta Jakarta Indonesia Dept. of Computer Engineering State University of Makassar Makassar Indonesia Dept. of Elec. and Comm. Engineering Okayama University Okayama Japan Dept. of Electrical Engineering State Polytechnic of Malang Malang Indonesia

Nowadays, computer-assisted learning (CAL) has experienced a significant increase in demand for self-learning purposes with the mechanism of automated assistance to facilitate assessments and provide inst.nt feedback to learners. Using automatic testing, test cases can be executed repeatedly, providing informative feedback to assist students in correcting the written code referring to the assignment specifications. This research experiences an automated testing mechanism in an Android programming self-learning system to serve several important functions for learning experience enhancement. By adopting the TDD method and utilizing automatic software testing tools, the testing process combines unit testing with JUnit and integration testing with Robolectric. It ensures the quality and functionality of the application under test based on learning assignment’s specifications. Utilizing the assertion methods realizes the code verification referring to expected behavior and potentially produces informative feedback that can be captured to provide automated assistance. Evaluation through the test code implementation in the Basic Application learning topic shows that automatic testing tools cover the automated assistance feature in the validations of project configuration and UI components.

关键词：

来源：评论

学校读者我要写书评

暂无评论

ASYMMETRIC CLEAN SEGMENTS-GUIDED SELF-SUPERVISED LEARNING FOR ROBUST SPEAKER VERIFICATION

arXiv

引用

arXiv 2023年

作者： Gan, Chong-Xin Mak, Man-Wai Lin, Weiwei Chien, Jen-Tzung Dept. of Electrical and Electronic Engineering The Hong Kong Polytechnic University Hong Kong Inst. of Electrical and Computer Engineering National Yang Ming Chiao Tung University Taiwan

Contrastive self-supervised learning (CSL) for speaker verification (SV) has drawn increasing interest recently due to its ability to exploit unlabeled data. Performing data augmentation on raw waveforms, such as adding noise or reverberation, plays a pivotal role in achieving promising results in SV. Data augmentation, however, demands meticulous calibration to ensure intact speaker-specific information, which is difficult to achieve without speaker labels. To address this issue, we introduce a novel framework by incorporating clean and augmented segments into the contrastive training pipeline. The clean segments are repurposed to pair with noisy segments to form additional positive and negative pairs. Moreover, the contrastive loss is weighted to increase the difference between the clean and augmented embeddings of different speakers. Experimental results on Voxceleb1 suggest that the proposed framework can achieve a remarkable 19% improvement over the conventional methods, and it surpasses many existing state-of-the-art techniques. Copyright © 2023, The Authors. All rights reserved.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

Contrastive Speaker Embedding With Sequential Disentanglement

arXiv

引用

arXiv 2023年

作者： Tu, Youzhi Mak, Man-Wai Chien, Jen-Tzung Dept. of Electrical and Electronic Engineering The Hong Kong Polytechnic University Hong Kong Inst. of Electrical and Computer Engineering National Yang Ming Chiao Tung University Taiwan

Contrastive speaker embedding assumes that the contrast between the positive and negative pairs of speech segments is attributed to speaker identity only. However, this assumption is incorrect because speech signals contain not only speaker identity but also linguistic content. In this paper, we propose a contrastive learning framework with sequential disentanglement to remove linguistic content by incorporating a disentangled sequential variational autoencoder (DSVAE) into the conventional SimCLR framework. The DSVAE aims to disentangle speaker factors from content factors in an embedding space so that only the speaker factors are used for constructing a contrastive loss objective. Because content factors have been removed from the contrastive learning, the resulting speaker embeddings will be content-invariant. Experimental results on VoxCeleb1-test show that the proposed method consistently outperforms SimCLR. This suggests that applying sequential disentanglement is beneficial to learning speaker-discriminative embeddings. Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Adverse Event Prevention on The Road System with Collaborative MEC

Adverse Event Prevention on The Road System with Collaborati...

引用

IEEE Conference on Vehicular technology (VTC)

作者： Ru-Jun Wang Han-Rong Lai Shih-Jui Wang Yu-Hsun Kuo Chih-Hang Wang Wen-Tsuen Chen De-Nian Yang Dept. of Computer Science National Tsing Hua University Hsinchu Taiwan Inst. of Communications Engineering National Tsing Hua University Hsinchu Taiwan Inst. of Information Science Academia Sinica Taiwan

The localization of Road Users (RUs) is an important issue in adverse event prevention due to the unreliable nature of GPS and the high cost of high-precision location acquisition sensors. In addition, previous research on adverse event prevention on roads has not taken into account RUs in blind spots at the same time. To address these issues, we investigate a collaboration system for heterogeneous RUs and Mobile Edge Computing (MEC) servers, called Collaborative Adverse Event Prevention system (CAEP) to efficiently alert RUs to potential adverse events and perceive the blind spot of the RUs. CAEP includes two AI-based functional modules, a localization module and a blind spot detection module, and an adverse event prevention algorithm. The localization module localizes each RU and the blind spot detection module detects the other RUs in the blind spot. The adverse event prevention algorithm jointly considers general road collision events and the event of a difference in radius between the inner wheels of a vehicle to completely include adverse events on the road. We implement CAEP in a real-world traffic environment with Jetson AGX Xavier devices and cameras to evaluate the performance. Our evaluation shows that CAEP provides RUs with sufficient preparation time to prevent adverse events and correctly detects the RUs in blind spots.

关键词：

来源：评论

学校读者我要写书评

暂无评论

UDA-CT: A General Framework for CT Image Standardization

UDA-CT: A General Framework for CT Image Standardization

引用

2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022

作者： Selim, Md Zhang, Jie Fei, Baowei Lewis, Matthew Zhang, Guo-Qiang Chen, Jin University of Kentucky Dept. of Computer Science LexingtonKY United States University of Kentucky Dept. of Radiology LexingtonKY United States University of Texas at Dallas Dept. of Bioengineering DallasTX United States UT Southwestern Medical Center Dept. of Radiology DallasTX United States University of Texas Health Science Center at Houston Dept. of Neurology HoustonTX United States University of Kentucky Inst. for Biomedical Informatics LexingtonKY United States

ISBN: (纸本)9781665468190

Large-scale CT image studies often suffer from a lack of homogeneity regarding radiomic characteristics due to the images acquired with scanners from different vendors or with different reconstruction algorithms. We propose a deep learning-based framework called UDA-CT to tackle the homogeneity issue by leveraging both paired and unpaired images. Using UDA-CT, the CT images can be standardized both from different acquisition protocols of the same scanner and CT images acquired using a similar protocol but scanners from different vendors. UDA-CT incorporates recent advances in deep learning including domain adaptation and adversarial augmentation. It includes a unique design for model training batch which integrates nonstandard images and their adversarial variations to enhance model generalizability. The experimental results show that UDA-CT significantly improves the performance of the cross-scanner image standardization by utilizing both paired and unpaired data. © 2022 IEEE.

关键词： Standardization

来源：评论

学校读者我要写书评

暂无评论

The Iterative Dichotomiser 3 Method Improves Soft Skills Training Decisions Using Survey Data

The Iterative Dichotomiser 3 Method Improves Soft Skills Tra...

引用

Machine Learning and Applied Network Technologies (ICMLANT), IEEE International Conference on

作者： Subir Gupta Quang-Trung Hoang Bhaswati Roy Subrata Chowdhury Sandip Mukherjee Aditya Kumar Pathak Diep Can Duc Dept. of CSE(AIML) Haldia Institute of Technology West Bengal India Faculty of Electrical and Electronic Engineering Phenikaa University Hanoi Vietnam Dept. of FMS Dr.B.C.Roy Engineering College Durgapur WB India Dept. of Computer Science and Machine Learning Sreenivasa Inst. of Technology and Management Studies Chittoor Andhra Pradesh India Dept. of IT Haldia Institute of Technology West Bengal India Electric Power University Hanoi Vietnam

ISBN: (数字)9798350389180

ISBN: (纸本)9798350389197

This investigation explores student expectations and perceptions regarding higher education inst.tutions, focusing on aligning professional programs with student career aspirations. The study evaluates how effectively academic inst.tutions understand and meet training needs essential for student success in professional fields. Students were categorized using the Iterative Dichotomiser 3 methodology to forecast their training requirements. Analytical tools such as the confusion matrix and the ROC curve assessed the precision of these predictions. The results show a specificity rate of 74.1 %, a sensitivity of 78.6%, and an overall accuracy of 76.7 % . The combined sensitivity and specificity score is 3.527, and the empirical area under the ROC curve is 0.76, suggesting the model's reasonable efficacy. The findings indicate a significant student demand for enhanced inst.uction in soft skills, underscoring the need for educational inst.tutions to adapt their curricula to better prepare students for professional challenges.

关键词： Training Surveys Technological innovation Sensitivity Accuracy Focusing Machine learning Sensitivity and specificity Iterative methods Decision trees

来源：评论

学校读者我要写书评

暂无评论

An Investigation of Running Load Comparisons of ARCore on Native Android and Unity for Outdoor Navigation System Using Smartphone

An Investigation of Running Load Comparisons of ARCore on Na...

引用

Vocational Education and Electrical engineering, International Conference on

作者： Komang Candra Brata Nobuo Funabiki Sritrusta Sukaridhoto Evianita Dewi Fajrianti Mustika Mentari Dept. of Information and Communication Systems Okayama University Okayama Japan Dept. of Informatics Engineering Universitas Brawijaya Malang Indonesia Dept. of Informatic and Computer Politeknik Elektronika Negeri Surabaya Surabaya Indonesia Dept. of Information Technology State Polytechnic of Malang Malang Indonesia

In recent years, the utilization of location-based augmented reality (LAR) has gained popularity across various domains, including tourism, education, gaming, and advertising. The advancements in AR technology have made it easier to develop immersive LAR applications. ARCore, a widely adopted AR development library, enables the creation of location-based AR apps on Android devices. However, the implementation of ARCore in real-time scenarios that demand significant computational resources, like outdoor pedestrian navigation, can vary depending on the chosen development environment. This research aims to explore and compare the running load performance of ARCore implementation in location-based AR applications developed with Unity and Android Native, which are two prevalent AR development platforms. The study involves building a prototype LAR application using both Unity and native Android, followed by comprehensive running load performance evaluations. The findings reveal that while both Unity and native Android apps can deliver compelling location-based AR experiences, native Android apps demonstrate superior performance and resource efficiency in outdoor scenarios, whereas Unity offers the advantage of multi-platform support and ease of development for complex 3D assets. These insights provide valuable guidance to developers in selecting the most suitable approach for developing location-based AR applications for navigation, aligning with their specific needs and requirements.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Framework for Smart & Secure Vehicle Intelligent Life Monitoring System

A Framework for Smart & Secure Vehicle Intelligent Life Moni...

引用

Artificial Intelligence in Education and Industry 4.0 (IDICAIEI), DMIHER International Conference on

作者： Ramesh Dhulipudi Srinivasa Rao Dangeti Venkateswarlu Pynam T Daniya Prasenjit Das Balajee Maram Department of CSE Godavari Institute of Engineering and Technology Rajahmundry Department of Information Technology Shri Vishnu Engineering College for women Dept. of IT JNTUGV CEV Vizianagaram Department of AI&ML(CSE) GMR Inst. Of Tech. Rajam AP Dept. Of Computer Science and Engineering Chandigarh University Mohali Punjab School of Computer Science and Artificial Intelligence SR University Warangal Telangana

In today's world, vehicle safety has become one of the most concerns for private. Nowadays, existing alarm systems of vehicles can be easily replaced as the necessary tools are equally available. While traveling safety is everyone's primary concern. This analysis explains the model of an efficient data model that can monitor and secure the condition of the vehicle / car. The use of IoT for tracking and monitoring is the components. Internet of Things (IoT) is significant for observing the vehicles, medical treatment and manufacturing approaches. By the uninterrupted growth of Internet of Things, many “things” are capable. In this framework, a new IoT based framework depended on Field Programmable Gate Array (FPGA) and System on Chip (SoC) that gives unified access for IoT is low as well as high-speed devices with extensibility and configurability. Hence, this model performs better accuracy and security.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Alzheimer's Disease Classification Using Genetic Data

Alzheimer's Disease Classification Using Genetic Data

引用

2021 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2021

作者： Khanal, Subash Chen, Jin Jacobs, Nathan Lin, Ai-Ling University of Kentucky Dept. of Computer Science LexingtonKY United States Inst. for Biomedical Informatics University of Kentucky LexingtonKY United States University of Missouri Dept. of Radiology ColumbiaMO United States

ISBN: (纸本)9781665401265

There has been a recent surge of interest in using genetic data to build ML-based accurate and interpretable disease classification models. In this line of research, we separately assess the potential of the peripheral blood gene expression data as well as the Single Nucleotide Polymorphism (SNP) data in building ML models for AD classification. We present a systematic approach on feature selection and ML model design using both types of genetic data provided by the Alzheimer's Disease Neuroimaging Initiatives (ADNI). Our two-step feature selection produced a curated list of important genes. In addition to these selected genetic features, to examine the role of non-genetic covariates, we included age and number of education years (EDU) as extra features. In the Control (CN) vs. AD classification, the best performing classifier, XGBoost, trained with gene expression features only and that with extra features included had Area Under Curve (AUC) of 0.64 and 0.65 respectively. However, AUC for the same task using SNP data only and that with extra features included was 0.56 and 0.64 respectively. The just above chance results of classifier trained with SNP features and the improvement when used along with additional covariates indicate low potential of SNP data in AD classification when used alone while also indicating the importance of non-genetic factors associated with AD. Nevertheless, with well above chance performance, gene expression features show great potential especially between groups of AD progression, i.e., CN vs. AD, CN vs. EMCI, EMCI vs. AD and LMCI vs. AD. The source code and manual are available at https://***/mvrl/ADNI_Genetics. © 2021 IEEE.

关键词： Feature Selection

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：