Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687030
Evangelos Paparidis, Konstantinos I. Kotis
Capturing knowledge in ontology-based AI applications may significantly propagate technical/statistical, cultural/social, cognitive/psychological, or other types of bias, to un-fair AI models and to their generated decisions. Biased ontologies (and consequently, knowledge graphs) engineered for intelligent surveillance applications can introduce technical barriers in fair capture of offenders, thus it must be researched as a first priority problem and a constant concern for explicit actions to be taken in the era of a more secure and fair world. In this paper we report preliminary research conducted on the novel topic of engineering fair ontologies and present first experiments with a prototype ontology and knowledge graph in the surveillance domain. Engineering fair ontologies is a quite new research topic, thus, the related work is at early stages. Having said that, in this paper we already highlight a recommended methodological approach for unbiasing ontologies, demonstrated in the surveillance domain, and we identify specific key research issues and challenges for further investigation by the ontology engineering community.
{"title":"Towards Engineering Fair Ontologies: Unbiasing a Surveillance Ontology","authors":"Evangelos Paparidis, Konstantinos I. Kotis","doi":"10.1109/PIC53636.2021.9687030","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687030","url":null,"abstract":"Capturing knowledge in ontology-based AI applications may significantly propagate technical/statistical, cultural/social, cognitive/psychological, or other types of bias, to un-fair AI models and to their generated decisions. Biased ontologies (and consequently, knowledge graphs) engineered for intelligent surveillance applications can introduce technical barriers in fair capture of offenders, thus it must be researched as a first priority problem and a constant concern for explicit actions to be taken in the era of a more secure and fair world. In this paper we report preliminary research conducted on the novel topic of engineering fair ontologies and present first experiments with a prototype ontology and knowledge graph in the surveillance domain. Engineering fair ontologies is a quite new research topic, thus, the related work is at early stages. Having said that, in this paper we already highlight a recommended methodological approach for unbiasing ontologies, demonstrated in the surveillance domain, and we identify specific key research issues and challenges for further investigation by the ontology engineering community.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132393225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687016
Siao-Ting Wang, Chenglie Du, Jinchao Chen, Zuo Zhao, Ying Zhang
With the development of embedded systems, people tend to abstract the capacity of an embedded equipment as a service in order to simplify the development, deployment, management and maintenance of the embedded software. By aggregating abilities of individual embedded devices via service composition, people can easily build a more reliable and efficient system. Although the service composition problem has been extensively studied in the field of Web, the adoptation of service composition technique to embedded devices stumps due to the resource limitation and platform heterogeneity of embedded systems. To address these problems, this paper builds a service composition engine for embedded systems that comprises three main works: First, this paper provides a uniform method to represent the embedded service composition problem. Second, this paper designs a compiling method based on topological sorting to convert the unified composition information into the service composition file that represents the way to implement the composite service. Third, this paper devises a bytecode virtual machine to execute the service composition file, and implements the composite service in a resource-friendly way. At last, a carefully devised experiment is conducted, and the result shows our devised engine provides a lightweight, reliable and well-performed way to realize the service composition technique on embedded devices.
{"title":"A Bytecode Service Composition Engine for Embedded Services","authors":"Siao-Ting Wang, Chenglie Du, Jinchao Chen, Zuo Zhao, Ying Zhang","doi":"10.1109/PIC53636.2021.9687016","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687016","url":null,"abstract":"With the development of embedded systems, people tend to abstract the capacity of an embedded equipment as a service in order to simplify the development, deployment, management and maintenance of the embedded software. By aggregating abilities of individual embedded devices via service composition, people can easily build a more reliable and efficient system. Although the service composition problem has been extensively studied in the field of Web, the adoptation of service composition technique to embedded devices stumps due to the resource limitation and platform heterogeneity of embedded systems. To address these problems, this paper builds a service composition engine for embedded systems that comprises three main works: First, this paper provides a uniform method to represent the embedded service composition problem. Second, this paper designs a compiling method based on topological sorting to convert the unified composition information into the service composition file that represents the way to implement the composite service. Third, this paper devises a bytecode virtual machine to execute the service composition file, and implements the composite service in a resource-friendly way. At last, a carefully devised experiment is conducted, and the result shows our devised engine provides a lightweight, reliable and well-performed way to realize the service composition technique on embedded devices.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134235266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687060
Man Zhang, Wei-dong Jin, Peng Tang, Liang Li
The dropper is one of the core equipment of high-speed railway catenary, and its working state affects the power supply stability of pantograph catenary system. In this paper, we propose an effective detection method of catenary dropper state based on target detection algorithm You Only Look Once (YOLOv3) and Out-of-Distribution Detector for Neural Networks (ODIN). This method uses YOLOv3 as dropper locating network to detect the dropper area in catenary. The designed dropper state classification model based on ODIN is trained by augmented dropper area images of various states, and then is applied to analyze the specific state of dropper area from the location area images which is output by dropper location network. The extensive experimental results show that YOLOv3 can accurately detect dropper. The ODIN can effectively eliminate the interference of locating errors on the classification of dropper state, and the detection performance of the dropper state classification model is significantly improved by data augmentation. On the testing set, the accuracy of dropper locating network is more than 94.1%, and the precision of dropper state classification model achieve 97.97%.
滴管是高速铁路接触网的核心设备之一，其工作状态直接影响受电弓接触网系统的供电稳定性。本文提出了一种基于目标检测算法You Only Look Once (YOLOv3)和out - distribution Detector for Neural Networks (ODIN)的有效悬链线滴管状态检测方法。该方法采用YOLOv3作为滴管定位网络，对悬链线上的滴管区域进行检测。设计的基于ODIN的滴管状态分类模型通过增强各种状态的滴管区域图像进行训练，然后利用滴管定位网络输出的位置区域图像对滴管区域的具体状态进行分析。大量的实验结果表明，YOLOv3可以准确地检测滴管。ODIN可以有效消除定位误差对液滴状态分类的干扰，通过数据增强，显著提高液滴状态分类模型的检测性能。在测试集上，液滴定位网络的精度达到94.1%以上，液滴状态分类模型的精度达到97.97%。
{"title":"A YOLOv3 and ODIN Based State Detection Method for High-speed Railway Catenary Dropper","authors":"Man Zhang, Wei-dong Jin, Peng Tang, Liang Li","doi":"10.1109/PIC53636.2021.9687060","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687060","url":null,"abstract":"The dropper is one of the core equipment of high-speed railway catenary, and its working state affects the power supply stability of pantograph catenary system. In this paper, we propose an effective detection method of catenary dropper state based on target detection algorithm You Only Look Once (YOLOv3) and Out-of-Distribution Detector for Neural Networks (ODIN). This method uses YOLOv3 as dropper locating network to detect the dropper area in catenary. The designed dropper state classification model based on ODIN is trained by augmented dropper area images of various states, and then is applied to analyze the specific state of dropper area from the location area images which is output by dropper location network. The extensive experimental results show that YOLOv3 can accurately detect dropper. The ODIN can effectively eliminate the interference of locating errors on the classification of dropper state, and the detection performance of the dropper state classification model is significantly improved by data augmentation. On the testing set, the accuracy of dropper locating network is more than 94.1%, and the precision of dropper state classification model achieve 97.97%.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134031405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687076
Keiju Hayashi, T. Migita, Norikazu Takahashi
Nonnegative Matrix Factorization (NMF) is an efficient dimensionality reduction method for nonnegative data. Recently, a distributed algorithm has been proposed for multiple agents in a network to execute the hierarchical alternating least squares algorithm, which is well known as a fast computation method for NMF. However, the average consensus algorithm used there requires each agent to store the entire history of the values of its variables until the complete average consensus is reached, which increases the memory usage and computational cost. In this paper, we propose to replace the complicated average consensus algorithm with a simple one, and show through simulations that this replacement does not degrade the quality of the result if the values of the hyper-parameters are properly chosen.
{"title":"Distributed HALS Algorithm for NMF based on Simple Average Consensus Algorithm","authors":"Keiju Hayashi, T. Migita, Norikazu Takahashi","doi":"10.1109/PIC53636.2021.9687076","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687076","url":null,"abstract":"Nonnegative Matrix Factorization (NMF) is an efficient dimensionality reduction method for nonnegative data. Recently, a distributed algorithm has been proposed for multiple agents in a network to execute the hierarchical alternating least squares algorithm, which is well known as a fast computation method for NMF. However, the average consensus algorithm used there requires each agent to store the entire history of the values of its variables until the complete average consensus is reached, which increases the memory usage and computational cost. In this paper, we propose to replace the complicated average consensus algorithm with a simple one, and show through simulations that this replacement does not degrade the quality of the result if the values of the hyper-parameters are properly chosen.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116566500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687064
Abdulgbar A. R. Farea, Chengliang Wang, Ebraheem Farea, Abdulfattah E. Ba Alawi
E-commerce, ticket booking, banking, and other web-based applications that deal with sensitive information, such as passwords, payment information, and financial information, are widespread. Some web developers may have different levels of understanding about securing an online application. The two vulnerabilities identified by the Open Web Application Security Project (OWASP) for its 2017 Top Ten List are SQL injection and Cross-site Scripting (XSS). Because of these two vulnerabilities, an attacker can take advantage of these flaws and launch harmful web-based actions. Many published articles concentrated on a binary classification for these attacks. This article developed a new approach for detecting SQL injection and XSS attacks using deep learning. SQL injection and XSS payloads datasets are combined into a single dataset. The word-embedding technique is utilized to convert the word’s text into a vector. Our model used BiLSTM to auto feature extraction, training, and testing the payloads dataset. BiLSTM classified the payloads into three classes: XSS, SQL injection attacks, and normal. The results showed great results in classifying payloads into three classes: XSS attacks, injection attacks, and non-malicious payloads. BiLSTM showed high performance reached 99.26% in terms of accuracy.
{"title":"Cross-Site Scripting (XSS) and SQL Injection Attacks Multi-classification Using Bidirectional LSTM Recurrent Neural Network","authors":"Abdulgbar A. R. Farea, Chengliang Wang, Ebraheem Farea, Abdulfattah E. Ba Alawi","doi":"10.1109/PIC53636.2021.9687064","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687064","url":null,"abstract":"E-commerce, ticket booking, banking, and other web-based applications that deal with sensitive information, such as passwords, payment information, and financial information, are widespread. Some web developers may have different levels of understanding about securing an online application. The two vulnerabilities identified by the Open Web Application Security Project (OWASP) for its 2017 Top Ten List are SQL injection and Cross-site Scripting (XSS). Because of these two vulnerabilities, an attacker can take advantage of these flaws and launch harmful web-based actions. Many published articles concentrated on a binary classification for these attacks. This article developed a new approach for detecting SQL injection and XSS attacks using deep learning. SQL injection and XSS payloads datasets are combined into a single dataset. The word-embedding technique is utilized to convert the word’s text into a vector. Our model used BiLSTM to auto feature extraction, training, and testing the payloads dataset. BiLSTM classified the payloads into three classes: XSS, SQL injection attacks, and normal. The results showed great results in classifying payloads into three classes: XSS attacks, injection attacks, and non-malicious payloads. BiLSTM showed high performance reached 99.26% in terms of accuracy.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"99 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127814483","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687022
Aihua Zhou, Li-Peng Zhu, Meng Xu, Sen Pan, Junfeng Qiao, Hongyan Qiu, Song Deng
To solve the problem of non-synchronization between enterprise application development and data development, this paper puts forward the concept of data middle platform, which combines the two data processing mechanisms of online analytical processing (OLAP) and online transaction processing (OLTP), so that faster and better data services can be provided to the foreground business. On this basis, this paper summarizes the research status of the related technologies of the data middle platform, including the architecture of the data middle platform and the key technologies of constructing the data middle platform. In-depth analysis of the business scale and business characteristics of OLTP and OLAP in various application scenarios, focusing on the technical difficulties in the application process of OLTP and OLAP in the application scenario. Finally, it summarizes the challenges faced by the basic research from three aspects: the construction of data middle platform, data quality assurance, and the application of mixed-thing analytical data management.
{"title":"Research on Mixed Transaction Analytical Data Management Oriented to Data Middle Platform","authors":"Aihua Zhou, Li-Peng Zhu, Meng Xu, Sen Pan, Junfeng Qiao, Hongyan Qiu, Song Deng","doi":"10.1109/PIC53636.2021.9687022","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687022","url":null,"abstract":"To solve the problem of non-synchronization between enterprise application development and data development, this paper puts forward the concept of data middle platform, which combines the two data processing mechanisms of online analytical processing (OLAP) and online transaction processing (OLTP), so that faster and better data services can be provided to the foreground business. On this basis, this paper summarizes the research status of the related technologies of the data middle platform, including the architecture of the data middle platform and the key technologies of constructing the data middle platform. In-depth analysis of the business scale and business characteristics of OLTP and OLAP in various application scenarios, focusing on the technical difficulties in the application process of OLTP and OLAP in the application scenario. Finally, it summarizes the challenges faced by the basic research from three aspects: the construction of data middle platform, data quality assurance, and the application of mixed-thing analytical data management.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128816207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687054
An Xu, Shaoyu Wang, Jingyi Fan, Xiujin Shi, Qiang Chen
Recently, many fully supervised deep learning based methods have been proposed for automatic cardiac segmentation. However, it is very expensive and time-consuming to annotate data for the task. In this paper, we present a novel dual attention based uncertainty-aware mean teacher semi-supervised framework (DA-UAMT) for cardiac image segmentation. The framework consists of a teacher model and a student model with the same structure and the student model learns from the teacher model by minimizing a segmentation loss generated from labeled images and a consistency loss generated from unlabeled images with respect to the targets of the teacher model. To enable the student model learn from more reliable targets, we introduce the Monte Carlo Dropout which estimates target uncertainty, and a novel dual attention mechanism which helps the network to focus on information in shape and channel dimension. To evaluate the proposed method, we conducted experiments on MICCAI 2017 Automated Cardiac Diagnosis Challenge (ACDC) dataset. Experiments show that our proposed DA-UAMT model is effective in utilizing unlabeled data to obtain considerably better segmentation of cardiac.
{"title":"Dual Attention Based Uncertainty-aware Mean Teacher Model for Semi-supervised Cardiac Image Segmentation","authors":"An Xu, Shaoyu Wang, Jingyi Fan, Xiujin Shi, Qiang Chen","doi":"10.1109/PIC53636.2021.9687054","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687054","url":null,"abstract":"Recently, many fully supervised deep learning based methods have been proposed for automatic cardiac segmentation. However, it is very expensive and time-consuming to annotate data for the task. In this paper, we present a novel dual attention based uncertainty-aware mean teacher semi-supervised framework (DA-UAMT) for cardiac image segmentation. The framework consists of a teacher model and a student model with the same structure and the student model learns from the teacher model by minimizing a segmentation loss generated from labeled images and a consistency loss generated from unlabeled images with respect to the targets of the teacher model. To enable the student model learn from more reliable targets, we introduce the Monte Carlo Dropout which estimates target uncertainty, and a novel dual attention mechanism which helps the network to focus on information in shape and channel dimension. To evaluate the proposed method, we conducted experiments on MICCAI 2017 Automated Cardiac Diagnosis Challenge (ACDC) dataset. Experiments show that our proposed DA-UAMT model is effective in utilizing unlabeled data to obtain considerably better segmentation of cardiac.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"349 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124301204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687068
Weihui Dai, Xinyue Li, Ziqing Xia, Jintian Zhou, Lijuan Song, H. Mao, Yan Kang
It is well known that the sense of smell has significant impacts on human moods, therefore olfactory effects have been widely applied to psychological adjustment as well as clinical treatment. Unlike other senses, smell works through the molecules of olfactory stimuli acting on the human nervous system to elicit psychological effects, which is difficult to be accurately described and measured. This makes the commonly used methods hardly applicable to olfactory affective computation. Through analysis of the neural mechanism of human emotions evoked by olfactory sense, this paper specifically designed an EEG experiment to obtain the neural activity data of olfactory stimuli, and compares the clustering characteristics of neural feature data with self-reported scores in PAD emotional space. Thereout, the LS-SVR estimator based on the feature parameters extracted from EEG signal data is proposed for olfactory affective computation. It shows better distinguishing performance and potential reliability than self-reported data, and thus provides an enlightening exploration of this issue.
{"title":"Olfactory Affective Computation Based on EEG Signal Data","authors":"Weihui Dai, Xinyue Li, Ziqing Xia, Jintian Zhou, Lijuan Song, H. Mao, Yan Kang","doi":"10.1109/PIC53636.2021.9687068","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687068","url":null,"abstract":"It is well known that the sense of smell has significant impacts on human moods, therefore olfactory effects have been widely applied to psychological adjustment as well as clinical treatment. Unlike other senses, smell works through the molecules of olfactory stimuli acting on the human nervous system to elicit psychological effects, which is difficult to be accurately described and measured. This makes the commonly used methods hardly applicable to olfactory affective computation. Through analysis of the neural mechanism of human emotions evoked by olfactory sense, this paper specifically designed an EEG experiment to obtain the neural activity data of olfactory stimuli, and compares the clustering characteristics of neural feature data with self-reported scores in PAD emotional space. Thereout, the LS-SVR estimator based on the feature parameters extracted from EEG signal data is proposed for olfactory affective computation. It shows better distinguishing performance and potential reliability than self-reported data, and thus provides an enlightening exploration of this issue.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126710481","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/PIC53636.2021.9687056
Lina Tong, Qianzhi Lin, Chuanlei Qin, Liang Peng
The number and location configuration methods of wearable sensors for human activity recognition (HAR) are analytically discussed. Based on the publicly available Daily and Sports Activities data set, a convolutional neural network (CNN) was built to recognize 19 kinds of daily and sports activities, and then the model was optimized for better performance. The results of numerous comparative experiments show that deep learning-based HAR is better than machine learning-based HAR in terms of accuracy, and its improvement in accuracy is not directly related to the increase of sensor quantity. Due to its strong capability of feature extraction, deep learning extracts not only activity-related features but also individual differences, therefore, the location with less individual randomness should be selected according to practical engineering. Moreover, the results are also influenced by the limb symmetry in the data set. Finally, the feasibility of achieving higher accuracy with fewer sensors is proved.
{"title":"A Comparison of Wearable Sensor Configuration Methods for Human Activity Recognition Using CNN","authors":"Lina Tong, Qianzhi Lin, Chuanlei Qin, Liang Peng","doi":"10.1109/PIC53636.2021.9687056","DOIUrl":"https://doi.org/10.1109/PIC53636.2021.9687056","url":null,"abstract":"The number and location configuration methods of wearable sensors for human activity recognition (HAR) are analytically discussed. Based on the publicly available Daily and Sports Activities data set, a convolutional neural network (CNN) was built to recognize 19 kinds of daily and sports activities, and then the model was optimized for better performance. The results of numerous comparative experiments show that deep learning-based HAR is better than machine learning-based HAR in terms of accuracy, and its improvement in accuracy is not directly related to the increase of sensor quantity. Due to its strong capability of feature extraction, deep learning extracts not only activity-related features but also individual differences, therefore, the location with less individual randomness should be selected according to practical engineering. Moreover, the results are also influenced by the limb symmetry in the data set. Finally, the feasibility of achieving higher accuracy with fewer sensors is proved.","PeriodicalId":297239,"journal":{"name":"2021 IEEE International Conference on Progress in Informatics and Computing (PIC)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121879973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：