Pub Date : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153233
Rifqi Nabila Zufar, D. Banjerdpongchai
Collision Avoidance System (CAS) is important for drone safety. CAS consists of three steps e.g., obstacle sensing, collision prediction, and collision avoidance. Collision prediction enables drones to gain information, process information, and estimate whether the object has a risk of collision. Convolution Neural Network (CNN) is one of the methods that can be employed for collision prediction. However, CNN is a method that needs a large data in the training. Dario Pedro et al. provided a dataset called the CoLANet dataset that consists of VDOs of collision drones. Subsequently, they proposed a new algorithm called Neural Network Pipeline which has a Convolution Neural Network (CNN) part to extract the feature from a couple of images. CNN extracts images by using MobileNetV2 as a pre-trained model. They chose MobileNetV2 due to training performance from another dataset. This paper aims to assess the performance of lightweight CNN models using the CoLANet dataset. The models will be trained on the Keras library with parameters of fewer than ten million. The models will be validated by Confusion Matrix and Receiver Operating Characteristics. In conclusion, we examine which pre-trained CNN model has the best performance and suggest ongoing work.
{"title":"Performance Comparison of Lightweight CNN Models for Drone Collision Avoidance Dataset","authors":"Rifqi Nabila Zufar, D. Banjerdpongchai","doi":"10.1109/ECTI-CON58255.2023.10153233","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153233","url":null,"abstract":"Collision Avoidance System (CAS) is important for drone safety. CAS consists of three steps e.g., obstacle sensing, collision prediction, and collision avoidance. Collision prediction enables drones to gain information, process information, and estimate whether the object has a risk of collision. Convolution Neural Network (CNN) is one of the methods that can be employed for collision prediction. However, CNN is a method that needs a large data in the training. Dario Pedro et al. provided a dataset called the CoLANet dataset that consists of VDOs of collision drones. Subsequently, they proposed a new algorithm called Neural Network Pipeline which has a Convolution Neural Network (CNN) part to extract the feature from a couple of images. CNN extracts images by using MobileNetV2 as a pre-trained model. They chose MobileNetV2 due to training performance from another dataset. This paper aims to assess the performance of lightweight CNN models using the CoLANet dataset. The models will be trained on the Keras library with parameters of fewer than ten million. The models will be validated by Confusion Matrix and Receiver Operating Characteristics. In conclusion, we examine which pre-trained CNN model has the best performance and suggest ongoing work.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117223216","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153142
Warakorn Luangluewut, P. Thiennviboon
Foreign exchange (Forex) currency trading is an attractive investment. Real profits from the Forex trading, like stock trading, come from differences between buying prices and selling prices, which can be recognized by exchange rate trends. Our goal is to predict a trend direction from the most recent set of exchange rates using a simple deep learning model. The Forex price time series is converted to an image which is used as an input of a convolution neural network. Uptrend/downtrend labels of Forex prices were recognized using Martingale strategy. The model was trained using EUR/USD 2013 and 2015 open price data. For testing performances, initial 26-minute samples of uptrend/ downtrend events from EUR/USD 2016-2018 and 2020 open price data were used and 93% accuracy is achieved. To show trading benefits, a simple trading algorithm was simulated using our model and common trend indicators for EUR/USD 2013/2015-2018/2020 and GBP/USD 2020. In most cases, the trading algorithm using our model gained profits about 6% – 422% higher than those using other techniques. Therefore, our proposed model, if used correctly, may provide substantial profits in Forex trading.
{"title":"Forex Price Trend Prediction using Convolutional Neural Network","authors":"Warakorn Luangluewut, P. Thiennviboon","doi":"10.1109/ECTI-CON58255.2023.10153142","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153142","url":null,"abstract":"Foreign exchange (Forex) currency trading is an attractive investment. Real profits from the Forex trading, like stock trading, come from differences between buying prices and selling prices, which can be recognized by exchange rate trends. Our goal is to predict a trend direction from the most recent set of exchange rates using a simple deep learning model. The Forex price time series is converted to an image which is used as an input of a convolution neural network. Uptrend/downtrend labels of Forex prices were recognized using Martingale strategy. The model was trained using EUR/USD 2013 and 2015 open price data. For testing performances, initial 26-minute samples of uptrend/ downtrend events from EUR/USD 2016-2018 and 2020 open price data were used and 93% accuracy is achieved. To show trading benefits, a simple trading algorithm was simulated using our model and common trend indicators for EUR/USD 2013/2015-2018/2020 and GBP/USD 2020. In most cases, the trading algorithm using our model gained profits about 6% – 422% higher than those using other techniques. Therefore, our proposed model, if used correctly, may provide substantial profits in Forex trading.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126957155","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153365
S. Plumworasawat, Napa Sae-Bae
This work studies the performance of the pretrained models to classify colorectal tissues and the effectiveness of tissue image classification models when it is applied across different datasets that may exhibit different color characteristics. In particular, the study is conducted on three pre-trained CNN models, ResNet50, VGG19, and EfficientNetB3, where they are used as to extract a feature vector of the image patch. Then the multi-layer perceptron-based neural network is used to perform the task of multiclass classification. The result shows that the best accuracy to classify the colorectal histopathological images was achieved by the ResNet model with 93.87% test accuracy on the Kather-texture2016 image dataset. And for the accuracy of the classification model when it is trained in the public dataset and applied to the local datasets, the normalized color image test dataset showed the best accuracy rate at 80.69% when compared to raw images and grayscale images among the local datasets. This result suggests that the pre-trained classification models could be useful for tissue image classification tasks across various laboratory sources and image color adjustment could help enhance the recognition performance even further of the image classification model. Therefore, consideration for image color correction would be needed to preserve the recognition performance of the model if the pre-trained tissue image classification model is to be used off-the-shelf across laboratories.
{"title":"Colorectal Tissue Image Classification Across Datasets","authors":"S. Plumworasawat, Napa Sae-Bae","doi":"10.1109/ECTI-CON58255.2023.10153365","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153365","url":null,"abstract":"This work studies the performance of the pretrained models to classify colorectal tissues and the effectiveness of tissue image classification models when it is applied across different datasets that may exhibit different color characteristics. In particular, the study is conducted on three pre-trained CNN models, ResNet50, VGG19, and EfficientNetB3, where they are used as to extract a feature vector of the image patch. Then the multi-layer perceptron-based neural network is used to perform the task of multiclass classification. The result shows that the best accuracy to classify the colorectal histopathological images was achieved by the ResNet model with 93.87% test accuracy on the Kather-texture2016 image dataset. And for the accuracy of the classification model when it is trained in the public dataset and applied to the local datasets, the normalized color image test dataset showed the best accuracy rate at 80.69% when compared to raw images and grayscale images among the local datasets. This result suggests that the pre-trained classification models could be useful for tissue image classification tasks across various laboratory sources and image color adjustment could help enhance the recognition performance even further of the image classification model. Therefore, consideration for image color correction would be needed to preserve the recognition performance of the model if the pre-trained tissue image classification model is to be used off-the-shelf across laboratories.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122689658","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153173
Pratchaya Pathomchaiwat, B. Neammanee, Jirawut Benjanarasut
The problem with connecting LCL grid connected inverters (LCL-GCI) to weak grids is that inverter operation is often unstable. And another problem is that it is difficult to analyze the stability of the LCL-GCI due to the nonlinearity of the system. Therefore, this paper presents a small signal model of LCL-GCI for analysis of inverter stability. The model is derived from the state space method to linearize the model around its operating point, used to carry out the poles of the LCL-GCI system to determine the system stability with a change of its parameters including weak grids conditions. A change of LCL-GCI and the grids parameters effects the stability of the system especially a weak grid conditions the grids impedance is increasing. The simulation shows the movement of system pole in s-plane demonstrates the system stability when LCL-GCI parameters and grid impedance are changed. Although the LCL-GCI is unstable under weak grids, the improvement parameter methods by increase LCL filter parameter and reduce the controller gain can bring the system stable again.
{"title":"Small Signal Model and Stability Analysis of LCL Grid Connected Inverter under Weak Grids","authors":"Pratchaya Pathomchaiwat, B. Neammanee, Jirawut Benjanarasut","doi":"10.1109/ECTI-CON58255.2023.10153173","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153173","url":null,"abstract":"The problem with connecting LCL grid connected inverters (LCL-GCI) to weak grids is that inverter operation is often unstable. And another problem is that it is difficult to analyze the stability of the LCL-GCI due to the nonlinearity of the system. Therefore, this paper presents a small signal model of LCL-GCI for analysis of inverter stability. The model is derived from the state space method to linearize the model around its operating point, used to carry out the poles of the LCL-GCI system to determine the system stability with a change of its parameters including weak grids conditions. A change of LCL-GCI and the grids parameters effects the stability of the system especially a weak grid conditions the grids impedance is increasing. The simulation shows the movement of system pole in s-plane demonstrates the system stability when LCL-GCI parameters and grid impedance are changed. Although the LCL-GCI is unstable under weak grids, the improvement parameter methods by increase LCL filter parameter and reduce the controller gain can bring the system stable again.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"190 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122858421","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153305
Vittayasak Rujivorakul
Health screenings can be challenging, particularly when dealing with a large number of participants, time constraints, and the risk of procedural infection. This study presents the implementation of a system that aims to assist in the registration of patients and label printing, as well as confirm screening results and notify participants of the results. When the system was applied to a graduation ceremony and four rounds of screening with over 9,321 people in total, it was found to be effective in reducing the average waiting time to 22.87 minutes, with 90% of participants waiting within an acceptable time of 30 minutes. It also reduces the time spent in the inspection area to no more than 5 minutes for all users of the system. It was also able to provide real-time updates. The system was able to improve the level of the health check-up process, providing a better experience for users and reducing the risk of infection in the screening area. It is a successful application of information systems to medical processes.
{"title":"Developing and Implementing a Real-Time Mass Health Screening System: MFU.Pass","authors":"Vittayasak Rujivorakul","doi":"10.1109/ECTI-CON58255.2023.10153305","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153305","url":null,"abstract":"Health screenings can be challenging, particularly when dealing with a large number of participants, time constraints, and the risk of procedural infection. This study presents the implementation of a system that aims to assist in the registration of patients and label printing, as well as confirm screening results and notify participants of the results. When the system was applied to a graduation ceremony and four rounds of screening with over 9,321 people in total, it was found to be effective in reducing the average waiting time to 22.87 minutes, with 90% of participants waiting within an acceptable time of 30 minutes. It also reduces the time spent in the inspection area to no more than 5 minutes for all users of the system. It was also able to provide real-time updates. The system was able to improve the level of the health check-up process, providing a better experience for users and reducing the risk of infection in the screening area. It is a successful application of information systems to medical processes.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114551824","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153298
Suphawit Julthochai, P. Khamsalee, R. Wongsan
This paper presents an experimental investigation into enhancing the performance of medium frequency small loop antennas for Through-the-Earth (TTE) communication at an operating frequency of 350 kHz. The proposed small loop antenna is designed to have a 1/50 wavelength for use in cave environments. The experimental setup comprises a transmitter system with a signal generator and an earth-grounded antenna, and a receiver system with a small loop antenna and a spectrum analyzer, which is tested by transmitting signals from the transmitter to the receiver at a surface testing location. To enhance the performance of the receiver antenna, we investigate the effect of increasing the number of antennas, from one to two and four elements, to determine the optimal array arrangement. The experimental results are evaluated based on received power and antenna impedance. Our findings indicate that using a 4x1 vertical array antenna with small loop antennas induces the highest currents from the earth and improves signal power reception by up to 11.6 dB compared to a single small loop antenna.
{"title":"An Experimental Study of Performance Enhancement of Medium-Frequency Small Loop Antennas for Through-the-Earth at 350 kHz","authors":"Suphawit Julthochai, P. Khamsalee, R. Wongsan","doi":"10.1109/ECTI-CON58255.2023.10153298","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153298","url":null,"abstract":"This paper presents an experimental investigation into enhancing the performance of medium frequency small loop antennas for Through-the-Earth (TTE) communication at an operating frequency of 350 kHz. The proposed small loop antenna is designed to have a 1/50 wavelength for use in cave environments. The experimental setup comprises a transmitter system with a signal generator and an earth-grounded antenna, and a receiver system with a small loop antenna and a spectrum analyzer, which is tested by transmitting signals from the transmitter to the receiver at a surface testing location. To enhance the performance of the receiver antenna, we investigate the effect of increasing the number of antennas, from one to two and four elements, to determine the optimal array arrangement. The experimental results are evaluated based on received power and antenna impedance. Our findings indicate that using a 4x1 vertical array antenna with small loop antennas induces the highest currents from the earth and improves signal power reception by up to 11.6 dB compared to a single small loop antenna.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122062240","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153316
Chanikarn Wanna, N. Chudpooti, W. Thaiwirot
The design of ultra-wideband antipodal Vivaldi antenna with parasitic patch is presented. The different shapes of parasitic patch i.e., elliptical shape, diamond shape, stepped shape which are placed on antenna aperture are proposed. The effect of these parasitic patches on reflection coefficient, gain, radiation pattern, front-to-back ratio and electric field distribution is discussed. The simulation results show that all parasitic patches can operate in frequency range from 3.1 to 10.6 GHz for UWB applications. Moreover, using these parasitic patches, the gain, radiation patterns, front-to-back ratio and electric field distribution are improved.
{"title":"Design of Ultra-Wideband Antipodal Vivaldi Antenna with different shapes of Parasitic Patch","authors":"Chanikarn Wanna, N. Chudpooti, W. Thaiwirot","doi":"10.1109/ECTI-CON58255.2023.10153316","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153316","url":null,"abstract":"The design of ultra-wideband antipodal Vivaldi antenna with parasitic patch is presented. The different shapes of parasitic patch i.e., elliptical shape, diamond shape, stepped shape which are placed on antenna aperture are proposed. The effect of these parasitic patches on reflection coefficient, gain, radiation pattern, front-to-back ratio and electric field distribution is discussed. The simulation results show that all parasitic patches can operate in frequency range from 3.1 to 10.6 GHz for UWB applications. Moreover, using these parasitic patches, the gain, radiation patterns, front-to-back ratio and electric field distribution are improved.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121375038","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153203
U. Prasatsap, N. Nernchad, S. Srita, T. Kaewchum, C. Termritthikun, S. Somkun
The single-phase grid-connected photovoltaic (PV) system was studied by simulating in MATLAB-Simulink program. This research aimed to study the influence of two parameters on the maximum power point tracking (MPPT) efficiency, e.g., a PV capacitor (CPV) and voltage step size. The maximum power point was tracked by using the perturbation and observation (P&O) algorithm with the PV voltage-loop and current-loop controller of the converter. The system was able to simulate a sudden increase or decline in irradiance. The result showed that the reducing voltage step size can track for better maximum power point of the PV array. The ripple voltage and ripple power have reduced values. The increased CPV value affects the overshoot, which has a reduced value and a faster response time for returning to a steady state. Therefore, the suitable CPV of 200 μF obtains the MPPT efficiency of 99.969% with a reduced overshoot and faster time response.
{"title":"Parameters Influence on MPPT Efficiency for Single-Phase Grid-Connected Photovoltaic System","authors":"U. Prasatsap, N. Nernchad, S. Srita, T. Kaewchum, C. Termritthikun, S. Somkun","doi":"10.1109/ECTI-CON58255.2023.10153203","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153203","url":null,"abstract":"The single-phase grid-connected photovoltaic (PV) system was studied by simulating in MATLAB-Simulink program. This research aimed to study the influence of two parameters on the maximum power point tracking (MPPT) efficiency, e.g., a PV capacitor (CPV) and voltage step size. The maximum power point was tracked by using the perturbation and observation (P&O) algorithm with the PV voltage-loop and current-loop controller of the converter. The system was able to simulate a sudden increase or decline in irradiance. The result showed that the reducing voltage step size can track for better maximum power point of the PV array. The ripple voltage and ripple power have reduced values. The increased CPV value affects the overshoot, which has a reduced value and a faster response time for returning to a steady state. Therefore, the suitable CPV of 200 μF obtains the MPPT efficiency of 99.969% with a reduced overshoot and faster time response.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128171010","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}
Recent advances in deep neural networks (DNNs) have significantly enhanced the capabilities of optical character recognition (OCR) technology, enabling its adoption to a wide range of real-world applications. Despite this success, DNN-based OCR is shown to be vulnerable to adversarial attacks, in which the adversary can influence the DNN model’s prediction by carefully manipulating input to the model. Prior work has demonstrated the security impacts of adversarial attacks on various OCR languages. However, to date, no studies have been conducted and evaluated on an OCR system tailored to the Thai language. To bridge this gap, this work presents a feasibility study of performing adversarial attacks on a specific Thai OCR application – Thai License Plate Recognition (LPR). Moreover, we propose a new type of adversarial attacks based on the semi-targeted scenario and show that this scenario is highly realistic in LPR applications. Our experimental results show the feasibility of our attacks as they can be performed on a commodity computer desktop with over 90% attack success rate.
{"title":"From Characters to Chaos: On the Feasibility of Attacking Thai OCR with Adversarial Examples","authors":"Chissanupong Jiamsuchon, Jakapan Suaboot, Norrathep Rattanavipanon","doi":"10.1109/ECTI-CON58255.2023.10153157","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153157","url":null,"abstract":"Recent advances in deep neural networks (DNNs) have significantly enhanced the capabilities of optical character recognition (OCR) technology, enabling its adoption to a wide range of real-world applications. Despite this success, DNN-based OCR is shown to be vulnerable to adversarial attacks, in which the adversary can influence the DNN model’s prediction by carefully manipulating input to the model. Prior work has demonstrated the security impacts of adversarial attacks on various OCR languages. However, to date, no studies have been conducted and evaluated on an OCR system tailored to the Thai language. To bridge this gap, this work presents a feasibility study of performing adversarial attacks on a specific Thai OCR application – Thai License Plate Recognition (LPR). Moreover, we propose a new type of adversarial attacks based on the semi-targeted scenario and show that this scenario is highly realistic in LPR applications. Our experimental results show the feasibility of our attacks as they can be performed on a commodity computer desktop with over 90% attack success rate.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"3 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115851220","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 : 2023-05-09DOI: 10.1109/ECTI-CON58255.2023.10153382
Nitas Suriyakan, Ekkawit Wangkanklang
A motor driver can be used with various types of motors, including AC motors, DC motors, and stepper motors. The function of a motor driver is to receive input signals from a control system and send output signals to the motor to achieve desired operations. This paper focused on controlling the DC motor using MOSFET hybrid with ATMEGA2560 microcontroller which generates pulse width modulation (PWM) signal to control the motor angular speed. As a result, the motor drive board design has obvious performance.
{"title":"Design and Invention of a Low-Cost Motor Driver for the wheeled Robot","authors":"Nitas Suriyakan, Ekkawit Wangkanklang","doi":"10.1109/ECTI-CON58255.2023.10153382","DOIUrl":"https://doi.org/10.1109/ECTI-CON58255.2023.10153382","url":null,"abstract":"A motor driver can be used with various types of motors, including AC motors, DC motors, and stepper motors. The function of a motor driver is to receive input signals from a control system and send output signals to the motor to achieve desired operations. This paper focused on controlling the DC motor using MOSFET hybrid with ATMEGA2560 microcontroller which generates pulse width modulation (PWM) signal to control the motor angular speed. As a result, the motor drive board design has obvious performance.","PeriodicalId":340768,"journal":{"name":"2023 20th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130905955","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：