Pub Date : 2019-10-01DOI: 10.1109/ICITEED.2019.8929973
Saowakhon Nookhao, Singha Chaveesuk
The intense competition of globalization and the advancement of technology cause the business sector to adjust and expand into digital business. In addition, the consumer behavior has also changed such as using a smartphone as a part of life and financial transactions. Using of electronic wallets is a new technology that has been popular for consumers for conducting financial transactions via smartphones. The objective of the research is to develop structural model of consumer trust to intention to use electronic wallet. It is also studied the factors that influence consumer trust and intention to use electronic wallet. It was found that consumer trust in electronic wallet is caused by 3 aspects; 1) Information quality which has four attributes (Accuracy, Completeness, Relevance, and Up-to-date) 2) System quality which has four attributes (Ease of use, Availability, Reliability, and User-friendly) and 3) Service quality which has three attributes (Assurance, Responsiveness, and Empathy). It was also found that consumer trust will affect the satisfaction and intention to use electronic wallet of consumers.
{"title":"The Consumer Trust Influencing Intention to Use Electronic Wallet in Thailand","authors":"Saowakhon Nookhao, Singha Chaveesuk","doi":"10.1109/ICITEED.2019.8929973","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929973","url":null,"abstract":"The intense competition of globalization and the advancement of technology cause the business sector to adjust and expand into digital business. In addition, the consumer behavior has also changed such as using a smartphone as a part of life and financial transactions. Using of electronic wallets is a new technology that has been popular for consumers for conducting financial transactions via smartphones. The objective of the research is to develop structural model of consumer trust to intention to use electronic wallet. It is also studied the factors that influence consumer trust and intention to use electronic wallet. It was found that consumer trust in electronic wallet is caused by 3 aspects; 1) Information quality which has four attributes (Accuracy, Completeness, Relevance, and Up-to-date) 2) System quality which has four attributes (Ease of use, Availability, Reliability, and User-friendly) and 3) Service quality which has three attributes (Assurance, Responsiveness, and Empathy). It was also found that consumer trust will affect the satisfaction and intention to use electronic wallet of consumers.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"34 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85527118","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929949
Shardha Porwal, S. Mittal
Ciphertext Policy Attribute Based Encryption techniques provide fine grained access control to securely share the data in the organizations where access rights of users vary according to their roles. We have noticed that various key delegation mechanisms are provided for CP-ABE schemes but no key delegation mechanism exists for CP-ABE with hidden access policy. In practical, users’ identity may be revealed from access policy in the organizations and unlimited further delegations may results in unauthorized data access. For maintaining the users’ anonymity, the access structure should be hidden and every user must be restricted for specified further delegations. In this work, we have presented a flexible secure key delegation mechanism for CP-ABE with hidden access structure. The proposed scheme enhances the capability of existing CP-ABE schemes by supporting flexible delegation, attribute revocation and user revocation with negligible enhancement in computational cost.
{"title":"A Flexible Secure Key Delegation Mechanism for CP-ABE with Hidden Access Structure","authors":"Shardha Porwal, S. Mittal","doi":"10.1109/ICITEED.2019.8929949","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929949","url":null,"abstract":"Ciphertext Policy Attribute Based Encryption techniques provide fine grained access control to securely share the data in the organizations where access rights of users vary according to their roles. We have noticed that various key delegation mechanisms are provided for CP-ABE schemes but no key delegation mechanism exists for CP-ABE with hidden access policy. In practical, users’ identity may be revealed from access policy in the organizations and unlimited further delegations may results in unauthorized data access. For maintaining the users’ anonymity, the access structure should be hidden and every user must be restricted for specified further delegations. In this work, we have presented a flexible secure key delegation mechanism for CP-ABE with hidden access structure. The proposed scheme enhances the capability of existing CP-ABE schemes by supporting flexible delegation, attribute revocation and user revocation with negligible enhancement in computational cost.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"28 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79272203","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929977
Sabari Vishnu Jayanthan Jaikrishnan, Orawan Chantarakasemchit, P. Meesad
Breast cancer is one of the most common cancers and is the major cause of cancer-related deaths in women worldwide. Breast cancer is a disease in which cells in the breast grow in a rapid state and out of control. Breast cancer can grow in different parts of the breast and may or may not spread outside the breast. If caught on early stage, the breast cancer can be cured before they spread but if the cancer cells have spread to other parts of the body, usually it is hard to cure. Early diagnosis of breast cancer might help increasing the life-span of cancer affected women. In this paper, a novel method for prediction of breast cancer, that enhances the accuracy using machine learning is proposed using six machine learning algorithms. The unbiased estimates of the algorithms are measured using k-fold cross-validation method. This proposed approach proves to increase the accuracy of traditional machine learning algorithms.
{"title":"A Breakup Machine Learning Approach for Breast Cancer Prediction","authors":"Sabari Vishnu Jayanthan Jaikrishnan, Orawan Chantarakasemchit, P. Meesad","doi":"10.1109/ICITEED.2019.8929977","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929977","url":null,"abstract":"Breast cancer is one of the most common cancers and is the major cause of cancer-related deaths in women worldwide. Breast cancer is a disease in which cells in the breast grow in a rapid state and out of control. Breast cancer can grow in different parts of the breast and may or may not spread outside the breast. If caught on early stage, the breast cancer can be cured before they spread but if the cancer cells have spread to other parts of the body, usually it is hard to cure. Early diagnosis of breast cancer might help increasing the life-span of cancer affected women. In this paper, a novel method for prediction of breast cancer, that enhances the accuracy using machine learning is proposed using six machine learning algorithms. The unbiased estimates of the algorithms are measured using k-fold cross-validation method. This proposed approach proves to increase the accuracy of traditional machine learning algorithms.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"22 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77479677","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929936
F. Aditya, L. M. Putranto, Sarjiya, Edwin Nugraha Putra, Eko Yudo Pramono, Marwah
This paper analyzes the solar (PV) power plant penetration on a unique condition. Some constraints bound the maximum penetration (allowed operation) of PV. One of those boundaries is uncertainty in load profile. The uncertainty of the load profile can give different maximum PV penetration. In this study, we determine the maximum PV penetration in two types of load profile. Those conditions are at the lowest demand and typical daily demand. The unit commitment result and net load analysis can be used to determine the maximum PV penetration. The result of this study shows that the maximum PV penetration at the lowest demand is lower than the daily demand. In 2021, the maximum penetration at the lowest demand in the Java–Bali system reaches 1,812 MWp. Meanwhile, the maximum PV penetration in the daily demand reaches 3,974 MWp.
{"title":"PV Operation on the Low Demand Condition in the Java–Bali System","authors":"F. Aditya, L. M. Putranto, Sarjiya, Edwin Nugraha Putra, Eko Yudo Pramono, Marwah","doi":"10.1109/ICITEED.2019.8929936","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929936","url":null,"abstract":"This paper analyzes the solar (PV) power plant penetration on a unique condition. Some constraints bound the maximum penetration (allowed operation) of PV. One of those boundaries is uncertainty in load profile. The uncertainty of the load profile can give different maximum PV penetration. In this study, we determine the maximum PV penetration in two types of load profile. Those conditions are at the lowest demand and typical daily demand. The unit commitment result and net load analysis can be used to determine the maximum PV penetration. The result of this study shows that the maximum PV penetration at the lowest demand is lower than the daily demand. In 2021, the maximum penetration at the lowest demand in the Java–Bali system reaches 1,812 MWp. Meanwhile, the maximum PV penetration in the daily demand reaches 3,974 MWp.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"7 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79017800","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}
Two tsunami events occurred in Indonesia in 2018 showed that earthquakes were not the only cause of tsunamis in Indonesia. Therefore, tsunami detection cannot only rely on earthquake sensors. In this case, HF radar is an alternative technology for sensing tsunamis by radio waves. In this paper, we examine the pros and cons of HF radar technology for tsunami detection by examining the mechanism of HF radars in detecting tsunamis as well as comparison to other tsunami sensors. The results of study show that HF radar can be technically functioned as an efficient tsunami sensor because of the wide radar coverage, real time, and continuous sensing. However, field testing and algorithm advancement are still being developed to improve the detection accuracy.
{"title":"Study on Detection Mechanism of HF Radar for Early Tsunami Detection and Comparison to Other Tsunami Sensors","authors":"Iswandi, Risanuri Hidayat, Budi Setiyanto, Sigit Budi Wibowo","doi":"10.1109/ICITEED.2019.8929984","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929984","url":null,"abstract":"Two tsunami events occurred in Indonesia in 2018 showed that earthquakes were not the only cause of tsunamis in Indonesia. Therefore, tsunami detection cannot only rely on earthquake sensors. In this case, HF radar is an alternative technology for sensing tsunamis by radio waves. In this paper, we examine the pros and cons of HF radar technology for tsunami detection by examining the mechanism of HF radars in detecting tsunamis as well as comparison to other tsunami sensors. The results of study show that HF radar can be technically functioned as an efficient tsunami sensor because of the wide radar coverage, real time, and continuous sensing. However, field testing and algorithm advancement are still being developed to improve the detection accuracy.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"46 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72974025","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929987
I. Imawati, F. D. Wijaya, Bambang Sugiyantoro
The transportation sector is the biggest consumer of gasoline and always increase in every year. Electric vehicle (EV) is an effort to reduce fossil energy usage. The electric vehicle being developed is a Golf cart. Golf cart EV uses the electric motor as its driving force. Types of electric motor often used is an induction motor. This research aims to design a low-voltage induction motor as the propulsion force with desired performance and operating characteristics. The design process uses ANSYS Maxwell RMxprt simulation software. The design proposed for Golf Cart EV uses three phases, 48V induction motor, with phase frequency of 50 Hz, output power of 3HP, and is using class C NEMA design. The design process is done by firstly determining the specifications of the motor, then calculating the main dimensions, windings, as well as the stator and rotor slots. When operating in its rated voltage, the induction motor has efficiency, power factor, and slip values consecutively are 86.07%, 0.82, and 3.62%. Motor characteristics can be seen from locked rotor torque, locked rotor current and breakdown torque from induction motor design information whose values consecutively are 222%, 605%, 308% of the rated value.
{"title":"Design and Simulation of Three Phase Squirrel Cage Induction Motor in Low Voltage System 48V 50Hz 3Hp for Electric Golf Cart","authors":"I. Imawati, F. D. Wijaya, Bambang Sugiyantoro","doi":"10.1109/ICITEED.2019.8929987","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929987","url":null,"abstract":"The transportation sector is the biggest consumer of gasoline and always increase in every year. Electric vehicle (EV) is an effort to reduce fossil energy usage. The electric vehicle being developed is a Golf cart. Golf cart EV uses the electric motor as its driving force. Types of electric motor often used is an induction motor. This research aims to design a low-voltage induction motor as the propulsion force with desired performance and operating characteristics. The design process uses ANSYS Maxwell RMxprt simulation software. The design proposed for Golf Cart EV uses three phases, 48V induction motor, with phase frequency of 50 Hz, output power of 3HP, and is using class C NEMA design. The design process is done by firstly determining the specifications of the motor, then calculating the main dimensions, windings, as well as the stator and rotor slots. When operating in its rated voltage, the induction motor has efficiency, power factor, and slip values consecutively are 86.07%, 0.82, and 3.62%. Motor characteristics can be seen from locked rotor torque, locked rotor current and breakdown torque from induction motor design information whose values consecutively are 222%, 605%, 308% of the rated value.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"94 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72704833","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929985
Ratcha Boonsuk, Chaitawatch Sudprasert, S. Supratid
This paper presents an investigation study on facial emotional expression recognition. Three linear-decision-boundaries classifiers: linear support vector classification (LSVC), linear discriminant analysis (LDA) and softmax (SM) techniques are utilized based on convolutional neural network (CNN) for efficient feature extraction, namely CNN-LSVC, CNN-LDA and CNN-SM respectively. Hyper-parameter tuning or selection needs the least effort for such three linear-decision-boundaries classifiers. In order to enhance recognition performance, particular image preprocessing: intensity transformation as well as image cropping technique are implemented before feeding input images into CNN feature extraction. Relying on 10-fold cross validation of 80%-20% training-testing CK+ dataset, above 90% average results of precision, recall, F1 scores and accuracy rates are yielded by all such three investigated methods. Confusion matrix is also determined for more-detail of results examination.
{"title":"An Investigation on Facial Emotional Expression Recognition Based on Linear-Decision-Boundaries Classifiers Using Convolutional Neural Network for Feature Extraction","authors":"Ratcha Boonsuk, Chaitawatch Sudprasert, S. Supratid","doi":"10.1109/ICITEED.2019.8929985","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929985","url":null,"abstract":"This paper presents an investigation study on facial emotional expression recognition. Three linear-decision-boundaries classifiers: linear support vector classification (LSVC), linear discriminant analysis (LDA) and softmax (SM) techniques are utilized based on convolutional neural network (CNN) for efficient feature extraction, namely CNN-LSVC, CNN-LDA and CNN-SM respectively. Hyper-parameter tuning or selection needs the least effort for such three linear-decision-boundaries classifiers. In order to enhance recognition performance, particular image preprocessing: intensity transformation as well as image cropping technique are implemented before feeding input images into CNN feature extraction. Relying on 10-fold cross validation of 80%-20% training-testing CK+ dataset, above 90% average results of precision, recall, F1 scores and accuracy rates are yielded by all such three investigated methods. Confusion matrix is also determined for more-detail of results examination.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"199 1","pages":"1-5"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80080904","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929965
Ade Pratama Suryo Kusumo, E. Firmansyah, F. D. Wijaya
Phase-locked loop (PLL) is a key algorithm in a grid-connected power converter. The PLL must have excellent performance but consume minimum resources when implemented in a microcontroller.As the phase detector in the PLL system, causing the second harmonic signal, a notch filter is applied to cancel the second and higher harmonic that caused in PLL system to get better output signal. A simple second-order notch filter was implemented on the microcontroller.The performance of implemented digital PLL will be measured through several experiments. The experiment results show that the proposed digital PLL performs well after tested with emulatedvoltage sag, distorted input waveforms, and phase-jump conditions.
{"title":"Performance of Digital Phase Locked Loop Based on Notch Filter in Industrial ARM-Cortex Microcontrollers","authors":"Ade Pratama Suryo Kusumo, E. Firmansyah, F. D. Wijaya","doi":"10.1109/ICITEED.2019.8929965","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929965","url":null,"abstract":"Phase-locked loop (PLL) is a key algorithm in a grid-connected power converter. The PLL must have excellent performance but consume minimum resources when implemented in a microcontroller.As the phase detector in the PLL system, causing the second harmonic signal, a notch filter is applied to cancel the second and higher harmonic that caused in PLL system to get better output signal. A simple second-order notch filter was implemented on the microcontroller.The performance of implemented digital PLL will be measured through several experiments. The experiment results show that the proposed digital PLL performs well after tested with emulatedvoltage sag, distorted input waveforms, and phase-jump conditions.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"81 1","pages":"1-4"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91228300","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8930000
H. Truong, C. Jeenanunta
In electrical power system, unit commitment is a fundamental problem for power system planning which aims to maintain electricity production continuously and respond the electricity consumption during every time interval of the scheduling horizon. For long-term operational planning, monthly unit commitment is especially concerned to optimally implement due to the adequate fuel provision for the available generating units over a month. In this study, optimizing the commitment with fuel limitation, daily peak load demand, and operational conditions for a month is considered. Accordingly, a mixed integer linear programming model is proposed with objective to ensure a sufficiency of fuel for the whole month, satisfying the total energy and covering the daily peak load demand with respecting to many operational constraints. The model is applied to a national level planning in Thailand power system which consists of various generator types including thermal, single-shaft gas turbine (SSGT), hydropower, combined-cycle-system gas turbine (CCGT), combined-cyclesystem steam turbine (CCST), and imported power which are purchased and supplied abroad. Simultaneously, development planning is examined and analyzed to reveal a generating unit status, producing output of each generator and fuel usage. Regarding the results, the decision makers can practically execute for scenario analysis of a variety of demand and fuel availability.
{"title":"Mixed Integer Linear Programming Model for Monthly Unit Commitment in the National Level Power System","authors":"H. Truong, C. Jeenanunta","doi":"10.1109/ICITEED.2019.8930000","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8930000","url":null,"abstract":"In electrical power system, unit commitment is a fundamental problem for power system planning which aims to maintain electricity production continuously and respond the electricity consumption during every time interval of the scheduling horizon. For long-term operational planning, monthly unit commitment is especially concerned to optimally implement due to the adequate fuel provision for the available generating units over a month. In this study, optimizing the commitment with fuel limitation, daily peak load demand, and operational conditions for a month is considered. Accordingly, a mixed integer linear programming model is proposed with objective to ensure a sufficiency of fuel for the whole month, satisfying the total energy and covering the daily peak load demand with respecting to many operational constraints. The model is applied to a national level planning in Thailand power system which consists of various generator types including thermal, single-shaft gas turbine (SSGT), hydropower, combined-cycle-system gas turbine (CCGT), combined-cyclesystem steam turbine (CCST), and imported power which are purchased and supplied abroad. Simultaneously, development planning is examined and analyzed to reveal a generating unit status, producing output of each generator and fuel usage. Regarding the results, the decision makers can practically execute for scenario analysis of a variety of demand and fuel availability.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"13 2 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83011973","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 : 2019-10-01DOI: 10.1109/ICITEED.2019.8929982
Vasinee Nussiri, P. Vateekul
Nowadays, many food images are posted on various social network platforms without identification labels. An automatic food categorization application would greatly help to identify and classify food categories. Food categorization is a complex problem since the number of category types can be more than one hundred. Many kinds of food are similar with only subtle differences in taste and presentation and this can lead to a problem called “finegrained issue”. Recently, a bilinear model was employed which showed good accuracy and generated excessive features to capture details among different food categories, albeit with limited performance. Diverse food categories require disparate sets of features. Here, an attention mechanism was applied to capture suitable features and specifically identify each food category. Furthermore, the performance of a bilinear backbone was also enhanced by applying Inception in correlation with Inception-ResNet-v2 and Inception-v3 networks. The experiment was conducted on the Wongnai dataset containing various images that were separated into 83 classes. Results showed that our attentional model outperformed the traditional bilinear model, with an average of 16% improvement showing 3% and 44% as min-max performance values, respectively.
{"title":"Food Image Categorization Using Attentional Bilinear Model","authors":"Vasinee Nussiri, P. Vateekul","doi":"10.1109/ICITEED.2019.8929982","DOIUrl":"https://doi.org/10.1109/ICITEED.2019.8929982","url":null,"abstract":"Nowadays, many food images are posted on various social network platforms without identification labels. An automatic food categorization application would greatly help to identify and classify food categories. Food categorization is a complex problem since the number of category types can be more than one hundred. Many kinds of food are similar with only subtle differences in taste and presentation and this can lead to a problem called “finegrained issue”. Recently, a bilinear model was employed which showed good accuracy and generated excessive features to capture details among different food categories, albeit with limited performance. Diverse food categories require disparate sets of features. Here, an attention mechanism was applied to capture suitable features and specifically identify each food category. Furthermore, the performance of a bilinear backbone was also enhanced by applying Inception in correlation with Inception-ResNet-v2 and Inception-v3 networks. The experiment was conducted on the Wongnai dataset containing various images that were separated into 83 classes. Results showed that our attentional model outperformed the traditional bilinear model, with an average of 16% improvement showing 3% and 44% as min-max performance values, respectively.","PeriodicalId":6598,"journal":{"name":"2019 11th International Conference on Information Technology and Electrical Engineering (ICITEE)","volume":"100 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80590183","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：