In the current revolution industry era, water is a must for everyone. For water to be available every time, there must be a mechanism for transporting water from one place to another. A water leak is a common problem when transporting water. Based on research in 2018 in Australia, it is estimated that 12% water is lost due to leakage. In some developing countries, the water loss rate can be higher. To find out the location of leakage, monitor about the habit of the system is needed. The common features used to locate leaks are water flow and pressure. Flow is more preferable feature to monitor due to its resistance to interference than pressure. For this purpose, an accurate and practical flow sensor is the Ultrasonic Water Flowmeter TUF-2000. This sensor can obtain a data stream containing water flow data in a pipe. We used the Local Outlier Factor (LOF) Algorithm to detect flow anomalies. The location of the anomaly also can be estimated based on sensor location. Our method can accurately detect sudden flow changes suspected of leak signatures. However, the challenge is the evolutionary flow changes due to small leaks, as this type of leak could result in the evolving outlier detection model.
{"title":"Leak Detection using Non-Intrusive Ultrasonic Water Flowmeter Sensor in Water Distribution Networks","authors":"A. M. Shiddiqi, Muhammad Baihaqi, Atar Babgei","doi":"10.1145/3575882.3575943","DOIUrl":"https://doi.org/10.1145/3575882.3575943","url":null,"abstract":"In the current revolution industry era, water is a must for everyone. For water to be available every time, there must be a mechanism for transporting water from one place to another. A water leak is a common problem when transporting water. Based on research in 2018 in Australia, it is estimated that 12% water is lost due to leakage. In some developing countries, the water loss rate can be higher. To find out the location of leakage, monitor about the habit of the system is needed. The common features used to locate leaks are water flow and pressure. Flow is more preferable feature to monitor due to its resistance to interference than pressure. For this purpose, an accurate and practical flow sensor is the Ultrasonic Water Flowmeter TUF-2000. This sensor can obtain a data stream containing water flow data in a pipe. We used the Local Outlier Factor (LOF) Algorithm to detect flow anomalies. The location of the anomaly also can be estimated based on sensor location. Our method can accurately detect sudden flow changes suspected of leak signatures. However, the challenge is the evolutionary flow changes due to small leaks, as this type of leak could result in the evolving outlier detection model.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115704032","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}
Face recognitions have been used for various activities, especially for online verification and security. Face recognition system is a simple biometric, however it is more vulnerable than other biometrics because human face is easy to be manipulated. Face Anti-Spoofing (FAS) system is one of methods for detecting attacks on face recognition system. In this paper, we propose a method for FAS by analyzing the image texture from OULU-NPU database using Local Binary Pattern (LBP) method with Convolutional Neural Network (CNN) as classifier. Our focus is on comparing optimizer on CNN and color representation on the data. The purpose is to find the best optimizer on CNN and the best color representation for FAS system. The FAS model is trained by half of the data from OULU-NPU database which is set in several color representations. The CNN is also set in several optimizers such as Adam, SGD, Adagrad, and RMSprop. The model that is trained in 50 epochs using HSV images with SGD optimizer achieves the best accuracy of 0.99 and area under curve (AUC) of 0.98 among 32 models. From the experiments, it was found that RMSprop optimizer was not suitable for this research.
{"title":"Comparison of Optimizer on Convolutional Neural Network and Color Representation on Data for Face Presentation Attack Detection","authors":"Nur Aisyah Nadiyah, A. Nugroho","doi":"10.1145/3575882.3575906","DOIUrl":"https://doi.org/10.1145/3575882.3575906","url":null,"abstract":"Face recognitions have been used for various activities, especially for online verification and security. Face recognition system is a simple biometric, however it is more vulnerable than other biometrics because human face is easy to be manipulated. Face Anti-Spoofing (FAS) system is one of methods for detecting attacks on face recognition system. In this paper, we propose a method for FAS by analyzing the image texture from OULU-NPU database using Local Binary Pattern (LBP) method with Convolutional Neural Network (CNN) as classifier. Our focus is on comparing optimizer on CNN and color representation on the data. The purpose is to find the best optimizer on CNN and the best color representation for FAS system. The FAS model is trained by half of the data from OULU-NPU database which is set in several color representations. The CNN is also set in several optimizers such as Adam, SGD, Adagrad, and RMSprop. The model that is trained in 50 epochs using HSV images with SGD optimizer achieves the best accuracy of 0.99 and area under curve (AUC) of 0.98 among 32 models. From the experiments, it was found that RMSprop optimizer was not suitable for this research.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127540194","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}
Arief Indriarto Haris, Rd. Angga Ferianda, A. Basuki
Today, one of the challenges and hot issues related to Electric Vehicles (EV) is communication security, especially in the exchange of sensitive information on EVs. Virtual Private Network (VPN) is present as an alternative for those who need a communication network with reliable connectivity and security. VPNs have different types of protocols, some of which are Point-to-Point Tunneling Protocol (PPTP) and Ethernet over IP (EoIP) combined with IPSec (EoIP/IPSec). The purpose of this study aimed to analyze and compare the Quality of Service (QoS) generated by PPTP and EoIP/IPSec, then see the potential for integration of the combination between EoIP and IPSec as an effort of securing communications on EVs. QoS measurements were carried out using the TIPHON standard whose parameters consist of delay, jitter, packet loss, and throughput, as well as additional parameters such as CPU consumption and network security review. Throughput testing on the VPN client side was also carried out. The results of this study indicate that the combination of EoIP and IPSec, both in terms of performance and security, is possible and has the potential to be integrated into EVs.
{"title":"QoS Analysis of Site-to-Site VPN and Its Integration Potential for Securing Communication on Electric Vehicles","authors":"Arief Indriarto Haris, Rd. Angga Ferianda, A. Basuki","doi":"10.1145/3575882.3575887","DOIUrl":"https://doi.org/10.1145/3575882.3575887","url":null,"abstract":"Today, one of the challenges and hot issues related to Electric Vehicles (EV) is communication security, especially in the exchange of sensitive information on EVs. Virtual Private Network (VPN) is present as an alternative for those who need a communication network with reliable connectivity and security. VPNs have different types of protocols, some of which are Point-to-Point Tunneling Protocol (PPTP) and Ethernet over IP (EoIP) combined with IPSec (EoIP/IPSec). The purpose of this study aimed to analyze and compare the Quality of Service (QoS) generated by PPTP and EoIP/IPSec, then see the potential for integration of the combination between EoIP and IPSec as an effort of securing communications on EVs. QoS measurements were carried out using the TIPHON standard whose parameters consist of delay, jitter, packet loss, and throughput, as well as additional parameters such as CPU consumption and network security review. Throughput testing on the VPN client side was also carried out. The results of this study indicate that the combination of EoIP and IPSec, both in terms of performance and security, is possible and has the potential to be integrated into EVs.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123791575","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}
Ihsan Nugraha, Inna Syafarina, I. Cartealy, Anis Hayati, Maulida Mazaya, S. Iryanto
National Research and Innovation Agency (BRIN)- Indonesia, hosts high performance computing (HPC) facilities to support research and innovation that need high computation resources. One example of a research area is bioinformatics. As sequencing technology advances, any lab with next generation sequencing (NGS) access can generate a huge amount of data in a very short time. However, the difficulties then have shifted to the data analysis step that follows. It usually requires significant computation resources, many specific tools that need to be chained together, and man resources that are familiar with command line syntax. In addition, the chaining of multiple tools into a comprehensive workflow is also difficult since one needs to understand both the computer system administration and biological information related to the problems they try to answer. These hinder the biologist to take advantage of sequencing technology for their research. In this technical report, we described our approaches to integrate Galaxy and BRIN HPC, to ease users to deploy their analysis workflow on BRIN HPC facility.
{"title":"Implementation of Workflow Engine on BRIN HPC Infrastructure","authors":"Ihsan Nugraha, Inna Syafarina, I. Cartealy, Anis Hayati, Maulida Mazaya, S. Iryanto","doi":"10.1145/3575882.3575958","DOIUrl":"https://doi.org/10.1145/3575882.3575958","url":null,"abstract":"National Research and Innovation Agency (BRIN)- Indonesia, hosts high performance computing (HPC) facilities to support research and innovation that need high computation resources. One example of a research area is bioinformatics. As sequencing technology advances, any lab with next generation sequencing (NGS) access can generate a huge amount of data in a very short time. However, the difficulties then have shifted to the data analysis step that follows. It usually requires significant computation resources, many specific tools that need to be chained together, and man resources that are familiar with command line syntax. In addition, the chaining of multiple tools into a comprehensive workflow is also difficult since one needs to understand both the computer system administration and biological information related to the problems they try to answer. These hinder the biologist to take advantage of sequencing technology for their research. In this technical report, we described our approaches to integrate Galaxy and BRIN HPC, to ease users to deploy their analysis workflow on BRIN HPC facility.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123845333","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}
Soedibyo, Avian Lukman Setya Budi, M. Ashari, D. Riawan
It is common that wind turbine system needs a voltage stability control algorithm to regulate the output so that the voltage remains stable so that changes that occur in the current can also be minimized. The lookup table is proven as one of the stability methods that is quite often used, especially in the electronics field, and from the previous results, voltage lookup table make the accurate result in voltage stability from the wind turbine. When the voltage set higher exceeding the wind turbine capability, the voltage will remain at maximum point obtained, giving higher power result. The purpose of this paper is to explain strategic policy model for the design of the voltage stabilizer for wind turbine using voltage lookup table. Remote areas, such as remote islands, often offer higher wind speed ratio per time. By the means, that the basic standard is quite different from the normal conditions in previous results. Strategic policy model had to be developed for the remote area condition. In the strategic policy model, there are several differences, such as current and voltage basis that increased slightly. From the result, it is proven that policy model is needed for better scheme and control due to correct calculation and placement.
{"title":"Strategic Policy Model of Wind Turbine Power Quality Control System Using Lookup Table in Remote Area","authors":"Soedibyo, Avian Lukman Setya Budi, M. Ashari, D. Riawan","doi":"10.1145/3575882.3575896","DOIUrl":"https://doi.org/10.1145/3575882.3575896","url":null,"abstract":"It is common that wind turbine system needs a voltage stability control algorithm to regulate the output so that the voltage remains stable so that changes that occur in the current can also be minimized. The lookup table is proven as one of the stability methods that is quite often used, especially in the electronics field, and from the previous results, voltage lookup table make the accurate result in voltage stability from the wind turbine. When the voltage set higher exceeding the wind turbine capability, the voltage will remain at maximum point obtained, giving higher power result. The purpose of this paper is to explain strategic policy model for the design of the voltage stabilizer for wind turbine using voltage lookup table. Remote areas, such as remote islands, often offer higher wind speed ratio per time. By the means, that the basic standard is quite different from the normal conditions in previous results. Strategic policy model had to be developed for the remote area condition. In the strategic policy model, there are several differences, such as current and voltage basis that increased slightly. From the result, it is proven that policy model is needed for better scheme and control due to correct calculation and placement.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123393991","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}
Mochamad Adityo Rachmadi, Galuh Prihantoro, T. Nugroho, D. Cahya, Heru Taufiqurrohman, Zaid Cahya
In designing an autonomous electric vehicle control system, a model representing the actual system could provide convenience to the research object because the design can focus only on the area that needs to be researched. However, it is necessary to know the dynamic characteristics of the scaled model, which is used as a reference in designing motion control in autonomous vehicles, especially when braking. Most researchers use low-cost mathematical modelling to test the dynamic reaction of the vehicle during braking. However, it might not fully show real-world responses, so using scaled models is expected to bridge the problem. Testing was carried out on a vehicle model with a scale of 1:10, which focused on stability during its braking. As a result, braking comfort is obtained with a deceleration value below 2.94 m/s2, a stopping time of more than 1.3 s and a stopping distance not exceeding from 200 cm. These results can be adopted as a guideline for the motion control design in the autonomous electric vehicles model.
{"title":"The Design of 1/10 Scale Model for Autonomous Electric Vehicle Dynamic Testing","authors":"Mochamad Adityo Rachmadi, Galuh Prihantoro, T. Nugroho, D. Cahya, Heru Taufiqurrohman, Zaid Cahya","doi":"10.1145/3575882.3575898","DOIUrl":"https://doi.org/10.1145/3575882.3575898","url":null,"abstract":"In designing an autonomous electric vehicle control system, a model representing the actual system could provide convenience to the research object because the design can focus only on the area that needs to be researched. However, it is necessary to know the dynamic characteristics of the scaled model, which is used as a reference in designing motion control in autonomous vehicles, especially when braking. Most researchers use low-cost mathematical modelling to test the dynamic reaction of the vehicle during braking. However, it might not fully show real-world responses, so using scaled models is expected to bridge the problem. Testing was carried out on a vehicle model with a scale of 1:10, which focused on stability during its braking. As a result, braking comfort is obtained with a deceleration value below 2.94 m/s2, a stopping time of more than 1.3 s and a stopping distance not exceeding from 200 cm. These results can be adopted as a guideline for the motion control design in the autonomous electric vehicles model.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125356994","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}
The existing data on the web is growing increasingly diverse and extensive. Unfortunately, most of the current data is not semantic web friendly. The relational database (RDB) is considered one of the most popular structured data models. This work aims to automatically migrate from an imperfect RDB to a resource description framework (RDF). This issue has been ignored by previous work. Imperfectness means the lack of constraints in RDB. We attempt to construct an algorithm to migrate the RDB (including imperfect ones) to RDF. Results and evaluation showed that the proposed algorithm has succeeded in finding the relationship in RDB and migrating the RDB to RDF.
{"title":"Automatic Migration From Imperfect Relational Database To Resource Description Framework","authors":"D. Wardani, Maulia Harjono","doi":"10.1145/3575882.3575954","DOIUrl":"https://doi.org/10.1145/3575882.3575954","url":null,"abstract":"The existing data on the web is growing increasingly diverse and extensive. Unfortunately, most of the current data is not semantic web friendly. The relational database (RDB) is considered one of the most popular structured data models. This work aims to automatically migrate from an imperfect RDB to a resource description framework (RDF). This issue has been ignored by previous work. Imperfectness means the lack of constraints in RDB. We attempt to construct an algorithm to migrate the RDB (including imperfect ones) to RDF. Results and evaluation showed that the proposed algorithm has succeeded in finding the relationship in RDB and migrating the RDB to RDF.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128690550","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}
Technological developments in the era of globalization, several companies are competing in the field of artificial intelligence by developing autonomous drive systems. Training and road segmentation testing in this study were carried out using deep learning with the U-Net architecture method. The advantage of this method over other methods is that U-Net retains the full context of the input image. The road segmentation algorithm is implemented using Python programming. The algorithm is then executed on a single board computer. The parameters sought are the first accuracy with a value of 91.65 %, the second precision with a value of 75.53 %, the third recall with a value of 93.19 %, the fourth F1-Score with a value of 82.99 %, and the last IOU with a value of 71.36 %. The live segmentation algorithm can still detect roads, but some scenes involving objects are not included in the road segmentation process. This is due to several factors such as confusing light conditions, blur, shadows, and colors. Live segmentation using RGB frame produces an average FPS value of 0.30, and without RGB produces an average FPS value of 1.82.
{"title":"Implementation of Road Segmentation Using U-Net Model on Single Board Computer","authors":"E. Prakasa, Dary Zhafran, Dwi Astharini","doi":"10.1145/3575882.3575928","DOIUrl":"https://doi.org/10.1145/3575882.3575928","url":null,"abstract":"Technological developments in the era of globalization, several companies are competing in the field of artificial intelligence by developing autonomous drive systems. Training and road segmentation testing in this study were carried out using deep learning with the U-Net architecture method. The advantage of this method over other methods is that U-Net retains the full context of the input image. The road segmentation algorithm is implemented using Python programming. The algorithm is then executed on a single board computer. The parameters sought are the first accuracy with a value of 91.65 %, the second precision with a value of 75.53 %, the third recall with a value of 93.19 %, the fourth F1-Score with a value of 82.99 %, and the last IOU with a value of 71.36 %. The live segmentation algorithm can still detect roads, but some scenes involving objects are not included in the road segmentation process. This is due to several factors such as confusing light conditions, blur, shadows, and colors. Live segmentation using RGB frame produces an average FPS value of 0.30, and without RGB produces an average FPS value of 1.82.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129966576","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}
M. Firdaus, D. K. Wibowo, M. Hamidah, Ryan Prasetya Utama, Mustika Fitriana Dewi, Mohammad Hamdani, L. Setianingrum, S. Rahardjo, Michael Andreas Purwoadi, Edhi Purnomo
The advantages of optical fiber sensors, including fiber bragg grating (FBG) sensors, e.g. greater sensitivity, immunity to electromagnetic interference (EMI) and radio frequency interference (RFI), versatility, reliability, compatibility to optical communication and telemetry, reduced cost, reduced size, and reduced weight have been very well known. In particular, the FBG sensors have been widely utilized in many fields such as in structural health monitoring, oil and gas, civil industry, medical and space equipment. In this paper we develop a model of FBG sensors embedded fiber optic communication (FOC) system by using OptiSystem v.18. The bit error rate (BER) and Q-factor analysis as results of the effect of two FBG sensors on telecommunication channels, in which the sensors are integrated into a single fiber optic communication (FOC) with 80 km of fiber optic length. The results show that even with 2 sensors embedded in FOC system, the min BER and max Q factor of the system are still in good conditions with values of 4.35x 10-7 and 4.87 respectively.
{"title":"The Effect of Fiber Bragg Grating (FBG) Sensors on Data Channel of Fiber Optic Communication (FOC) System","authors":"M. Firdaus, D. K. Wibowo, M. Hamidah, Ryan Prasetya Utama, Mustika Fitriana Dewi, Mohammad Hamdani, L. Setianingrum, S. Rahardjo, Michael Andreas Purwoadi, Edhi Purnomo","doi":"10.1145/3575882.3575890","DOIUrl":"https://doi.org/10.1145/3575882.3575890","url":null,"abstract":"The advantages of optical fiber sensors, including fiber bragg grating (FBG) sensors, e.g. greater sensitivity, immunity to electromagnetic interference (EMI) and radio frequency interference (RFI), versatility, reliability, compatibility to optical communication and telemetry, reduced cost, reduced size, and reduced weight have been very well known. In particular, the FBG sensors have been widely utilized in many fields such as in structural health monitoring, oil and gas, civil industry, medical and space equipment. In this paper we develop a model of FBG sensors embedded fiber optic communication (FOC) system by using OptiSystem v.18. The bit error rate (BER) and Q-factor analysis as results of the effect of two FBG sensors on telecommunication channels, in which the sensors are integrated into a single fiber optic communication (FOC) with 80 km of fiber optic length. The results show that even with 2 sensors embedded in FOC system, the min BER and max Q factor of the system are still in good conditions with values of 4.35x 10-7 and 4.87 respectively.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115812151","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}
Budiarianto Suyo Kusumo, Siwi Oktaviana, W. Sulandari, A. Heryana, R. S. Yuwana, Endang Suryawati, A. R. Yuliani, H. Pardede
The Increasing of road mobility triggers the increasing of the number of traffic accidents. One of the main factors of the accidents is human errors which are heavily influenced by the driver conditions. Fatigue, drowsiness, and loss of concentration are among the common driver conditions that could cause traffic accident in addition to high-speed driving behavior. This could be minimized if early warning systems of driver conditions existed. This research aims to develop an early detection system for driver conditions using Convolutional Neural Network (CNN) method. Here, we investigate the effect of the depth of CNN and other hyper-parameters and observe their performance. We used eye movements and mouth conditions to be an indicator driver conditions. We evaluate the method using public dataset that contains image data of drivers on the highway in a state of yawning, not yawning, eyes open, and eyes closed. The experiment showed the best parameters with a learning rate of 0.001 and an epoch of 100. The resulting accuracy reached 99.31%.
{"title":"Detection of Driver Drowsiness Based on Eye and Mouth Movements Using Convolutional Neural Networks","authors":"Budiarianto Suyo Kusumo, Siwi Oktaviana, W. Sulandari, A. Heryana, R. S. Yuwana, Endang Suryawati, A. R. Yuliani, H. Pardede","doi":"10.1145/3575882.3575909","DOIUrl":"https://doi.org/10.1145/3575882.3575909","url":null,"abstract":"The Increasing of road mobility triggers the increasing of the number of traffic accidents. One of the main factors of the accidents is human errors which are heavily influenced by the driver conditions. Fatigue, drowsiness, and loss of concentration are among the common driver conditions that could cause traffic accident in addition to high-speed driving behavior. This could be minimized if early warning systems of driver conditions existed. This research aims to develop an early detection system for driver conditions using Convolutional Neural Network (CNN) method. Here, we investigate the effect of the depth of CNN and other hyper-parameters and observe their performance. We used eye movements and mouth conditions to be an indicator driver conditions. We evaluate the method using public dataset that contains image data of drivers on the highway in a state of yawning, not yawning, eyes open, and eyes closed. The experiment showed the best parameters with a learning rate of 0.001 and an epoch of 100. The resulting accuracy reached 99.31%.","PeriodicalId":367340,"journal":{"name":"Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128406982","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：