Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604652
S. Eyupoglu
In recent years, electronic textiles are growing, and light textile materials have started to replace heavy weight metallic materials. The conversion from traditional electronics into wearable electronics has acquired a significant interest owing to the improvement in material science. Furthermore, electronic textiles have attracted attention because of various applications in electromagnetic shielding effect, biomedical devices, energy conversion and storage. Electronic textiles can be defined as textile material having electrical properties. Electrical properties can be gained to textile materials with a variety of techniques such as chemical metallization, utilization of electrical conductive substrates in the polymer before fiber extrusion, weaving metallic wires into fabric, impregnation of conductive polymers to textile materials, and coating of conductive polymers to textile materials. In this study, wearable electronic textiles and production methods were investigated.
{"title":"Electronic Textiles: A Review","authors":"S. Eyupoglu","doi":"10.1109/ISMSIT52890.2021.9604652","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604652","url":null,"abstract":"In recent years, electronic textiles are growing, and light textile materials have started to replace heavy weight metallic materials. The conversion from traditional electronics into wearable electronics has acquired a significant interest owing to the improvement in material science. Furthermore, electronic textiles have attracted attention because of various applications in electromagnetic shielding effect, biomedical devices, energy conversion and storage. Electronic textiles can be defined as textile material having electrical properties. Electrical properties can be gained to textile materials with a variety of techniques such as chemical metallization, utilization of electrical conductive substrates in the polymer before fiber extrusion, weaving metallic wires into fabric, impregnation of conductive polymers to textile materials, and coating of conductive polymers to textile materials. In this study, wearable electronic textiles and production methods were investigated.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133891214","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604610
Ufuk Korkmaz, Cem Sanga, Deniz Türkpençe
The dissipative model of quantum computation is proven to be equivalent to its circuit model. Particularly, multi-qubit gates require time-dependent control with optimized parameters for some specific problems. One such problem is the simulation of a quantum version of a perceptron that classifies quantum information as binary using the framework of open quantum systems. In this scheme, a probe qubit is in contact with multiple, distinct quantum information-bearing environments and returns a binary decision depending on its amplitude parameter in its steady state. We refer to these environments as quantum information reservoirs. We choose a standard quantum collisional model in which the reservoir parameters can be defined in detail. In this study, we present the analytical results of the proposed classifier with an application to the superconductor quantum circuits for a single information reservoir. We exploit the additivity of quantum dynamic maps for dissipative processes in the weak coupling regime where optimized time-dependent control is not required to achieve the classification result. We show that the current state-of-the-art for superconducting circuits allows for the physical implementation of dissipative quantum information processing in the presence of information reservoirs with realistic parameters.
{"title":"Mimicking an Information Reservoir by Superconducting Quantum Circuits","authors":"Ufuk Korkmaz, Cem Sanga, Deniz Türkpençe","doi":"10.1109/ISMSIT52890.2021.9604610","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604610","url":null,"abstract":"The dissipative model of quantum computation is proven to be equivalent to its circuit model. Particularly, multi-qubit gates require time-dependent control with optimized parameters for some specific problems. One such problem is the simulation of a quantum version of a perceptron that classifies quantum information as binary using the framework of open quantum systems. In this scheme, a probe qubit is in contact with multiple, distinct quantum information-bearing environments and returns a binary decision depending on its amplitude parameter in its steady state. We refer to these environments as quantum information reservoirs. We choose a standard quantum collisional model in which the reservoir parameters can be defined in detail. In this study, we present the analytical results of the proposed classifier with an application to the superconductor quantum circuits for a single information reservoir. We exploit the additivity of quantum dynamic maps for dissipative processes in the weak coupling regime where optimized time-dependent control is not required to achieve the classification result. We show that the current state-of-the-art for superconducting circuits allows for the physical implementation of dissipative quantum information processing in the presence of information reservoirs with realistic parameters.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114758435","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604571
Görkem Melih Kıranel, B. Ataşlar-Ayyıldız
In this work, a PID controller tuned with the multi-objective optimization is used to realize high precision tracking control of a magnetic levitation system. For the multi-objective optimization, a cost function based on the system dynamic response criteria is proposed. The weights of the multi-objective cost function are calculated by three different optimization algorithms: PSO, ABC, and GWO. Time domain performance of the system response obtained by the simulations for these three controller are compared. In addition, the performance of the obtained controllers is examined for a different reference signal and even under the disturbance.
{"title":"Optimal Tuning of PID Parameters for Magnetic Ball Levitation System Using Multi-objective Evolutionary Algorithms","authors":"Görkem Melih Kıranel, B. Ataşlar-Ayyıldız","doi":"10.1109/ISMSIT52890.2021.9604571","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604571","url":null,"abstract":"In this work, a PID controller tuned with the multi-objective optimization is used to realize high precision tracking control of a magnetic levitation system. For the multi-objective optimization, a cost function based on the system dynamic response criteria is proposed. The weights of the multi-objective cost function are calculated by three different optimization algorithms: PSO, ABC, and GWO. Time domain performance of the system response obtained by the simulations for these three controller are compared. In addition, the performance of the obtained controllers is examined for a different reference signal and even under the disturbance.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115851855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604737
Ali Nuri Şeker, Hüseyin Doğan, Muhammet Üsame Öziç
In this Study a novel method has been used for extracting the background from given images. Different from the existing approaches, a Convolutional Neural Network (CNN) Autoencoder (AE) has been trained with frames produced from the same stationary camera source, paired with random frames from the same pool for each sample as a label. A little over 4000 RGB images with the dimensions of 640x480 has been used for training and around 450 of them was used for testing. The mentioned model has 4 convolutional layers each in encoder and decoder sections. The training was conducted for 500 epochs and the value of epoch loss went down to 2.13x10-3 and 2.38x10-3 for training and validation respectively. After the training of the model, generated background samples were subtracted from the input images and was turned into a binary image using two different segmentation methods: HSV Thresholding and OTSU. To use as the ground truth, test images were hand labeled. Mentioned approach had an F1-score of 62.36% for HSV Thresholding and 69.63% for OTSU methods.
{"title":"A Novel Approach on Unsupervised Dynamic Background Extraction Using Autoencoders","authors":"Ali Nuri Şeker, Hüseyin Doğan, Muhammet Üsame Öziç","doi":"10.1109/ISMSIT52890.2021.9604737","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604737","url":null,"abstract":"In this Study a novel method has been used for extracting the background from given images. Different from the existing approaches, a Convolutional Neural Network (CNN) Autoencoder (AE) has been trained with frames produced from the same stationary camera source, paired with random frames from the same pool for each sample as a label. A little over 4000 RGB images with the dimensions of 640x480 has been used for training and around 450 of them was used for testing. The mentioned model has 4 convolutional layers each in encoder and decoder sections. The training was conducted for 500 epochs and the value of epoch loss went down to 2.13x10-3 and 2.38x10-3 for training and validation respectively. After the training of the model, generated background samples were subtracted from the input images and was turned into a binary image using two different segmentation methods: HSV Thresholding and OTSU. To use as the ground truth, test images were hand labeled. Mentioned approach had an F1-score of 62.36% for HSV Thresholding and 69.63% for OTSU methods.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"51 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123559589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604577
Noor Qasim Atiyah Alsaedi, Muhammet Tahir Guneser
In this study we proposed capacitive power transfer (CPT) plate coupler of four-plates using for large distance airgap. We evaluated the effect of square and circular plates. In this study we compared capacitive horizontal plates and vertical plates and find the outcome of different size to maintain the coupling of two plates side. This model coupler circuit is approaching, making an allowance for all coupling capacitors. Highly power transfer (high plate voltage) and resonating coupler have been supplied by using the Double-sided LCLC compensated topology of circuit. The coupler circuit model easy to design the compensation circuit parameters. Analysis of Finite-element is used the coupling to simulate capacitance and choose the sizes of coupler. performance of circuit is simulated in ANSYS to proposal the precise parameter values.
{"title":"Geometry Optimization of Wireless Capacitive Power Transfer Plates for Electrical Vehicle","authors":"Noor Qasim Atiyah Alsaedi, Muhammet Tahir Guneser","doi":"10.1109/ISMSIT52890.2021.9604577","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604577","url":null,"abstract":"In this study we proposed capacitive power transfer (CPT) plate coupler of four-plates using for large distance airgap. We evaluated the effect of square and circular plates. In this study we compared capacitive horizontal plates and vertical plates and find the outcome of different size to maintain the coupling of two plates side. This model coupler circuit is approaching, making an allowance for all coupling capacitors. Highly power transfer (high plate voltage) and resonating coupler have been supplied by using the Double-sided LCLC compensated topology of circuit. The coupler circuit model easy to design the compensation circuit parameters. Analysis of Finite-element is used the coupling to simulate capacitance and choose the sizes of coupler. performance of circuit is simulated in ANSYS to proposal the precise parameter values.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125538175","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}
One of the rich data fields is the biomedical realm. There are detailed biomedical records, ranging from clinical conditions to different types of biochemical data and imaging equipment outputs. However, the manual extraction and transformation of biomedical patterns into mechanically comprehensible information is a cumbersome challenge since the biomedical field includes broad, dynamic, and complex knowledge. The main focus of this study is to analyze some of the available data mining patterns for the classification of medical cases with the systematic literature review method. Its emphasis is on studying techniques that are commonly used for the prognosis, classification, prediction, and treatment-related to recurrent and significant diseases like cancer, hepatitis, and cardiac diseases. Data mining can enhance healthcare choices and patient survival time. The researchers of this study hope that this research provides information on the data mining classification algorithms used to study medical cases that are not even mentioned in this study.
{"title":"Data Mining Techniques for the Classification of Medical Cases: A Survey","authors":"Oluwaseun Priscilla Olawale, Fezile Ozdamli, Kamil Dimililer","doi":"10.1109/ISMSIT52890.2021.9604724","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604724","url":null,"abstract":"One of the rich data fields is the biomedical realm. There are detailed biomedical records, ranging from clinical conditions to different types of biochemical data and imaging equipment outputs. However, the manual extraction and transformation of biomedical patterns into mechanically comprehensible information is a cumbersome challenge since the biomedical field includes broad, dynamic, and complex knowledge. The main focus of this study is to analyze some of the available data mining patterns for the classification of medical cases with the systematic literature review method. Its emphasis is on studying techniques that are commonly used for the prognosis, classification, prediction, and treatment-related to recurrent and significant diseases like cancer, hepatitis, and cardiac diseases. Data mining can enhance healthcare choices and patient survival time. The researchers of this study hope that this research provides information on the data mining classification algorithms used to study medical cases that are not even mentioned in this study.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126172050","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604584
Beyza Nur Usta, Berranur Tepeyurt, E. Karakulak
Bio-potential signal generators are widely used for education, testing of devices and scientific purposes. Due to their high prices, these generators cannot be supplied in sufficient numbers. In this study, an ECG generator that generates a simple synthetic ECG and is inexpensive is proposed. The generator obtains the ECG signal with a simple ECG model and the internal Pulse Width Modulation (PWM) port of the microcontroller is used to obtain the analog signal. In this way, an ECG signal generator can be obtained with less electronic material. The circuit is built and the experimental results of the circuit are given.
{"title":"Simple Synthetic ECG Generation via PWM Output of Microcontroller","authors":"Beyza Nur Usta, Berranur Tepeyurt, E. Karakulak","doi":"10.1109/ISMSIT52890.2021.9604584","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604584","url":null,"abstract":"Bio-potential signal generators are widely used for education, testing of devices and scientific purposes. Due to their high prices, these generators cannot be supplied in sufficient numbers. In this study, an ECG generator that generates a simple synthetic ECG and is inexpensive is proposed. The generator obtains the ECG signal with a simple ECG model and the internal Pulse Width Modulation (PWM) port of the microcontroller is used to obtain the analog signal. In this way, an ECG signal generator can be obtained with less electronic material. The circuit is built and the experimental results of the circuit are given.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"99 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125868690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604534
S. Widiyanto, D. T. Wardani, Singgih Wisnu Pranata
Tomato is one of the cultivations that is often used for gardening. Tomato also has high demands on the market as it’s used for daily needs and occurred for many cuisines. Tomato comes in several colors such as red, orange, and green. Their color could tell their maturity levels too. Tomato grows in several quantities even only on one branch. So as the technologies grow, the computer also could be trained to understand what tomato is and how does it look like. Using computer vision, the computer could tell tomatoes according to their color. For this study, the computer will be trained using Faster R-CNN models to recognize the tomato maturity as Faster R-CNN known support for the image classification and object detection. The accuracy for classification in validation stage about 98,70% in average. For the object detection the model has confidentiality about 96,20% to detect the tomato maturity.
{"title":"Image-Based Tomato Maturity Classification and Detection Using Faster R-CNN Method","authors":"S. Widiyanto, D. T. Wardani, Singgih Wisnu Pranata","doi":"10.1109/ISMSIT52890.2021.9604534","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604534","url":null,"abstract":"Tomato is one of the cultivations that is often used for gardening. Tomato also has high demands on the market as it’s used for daily needs and occurred for many cuisines. Tomato comes in several colors such as red, orange, and green. Their color could tell their maturity levels too. Tomato grows in several quantities even only on one branch. So as the technologies grow, the computer also could be trained to understand what tomato is and how does it look like. Using computer vision, the computer could tell tomatoes according to their color. For this study, the computer will be trained using Faster R-CNN models to recognize the tomato maturity as Faster R-CNN known support for the image classification and object detection. The accuracy for classification in validation stage about 98,70% in average. For the object detection the model has confidentiality about 96,20% to detect the tomato maturity.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126020966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604539
K. O. Canbek, Y. Oniz
In this study, a Type-2 Fuzzy Logic Controller that directly generates the necessary control signals is used to achieve the trajectory tracking control of an unmanned aerial vehicle (UAV). The proposed control scheme takes the quadcopter’s divergence from the given trajectory and its derivative as inputs. A type-1 fuzzy logic controller is also designed and implemented on the quadcopter to assess the performance of the type-2 controller. Real-time flight experiments are conducted for a skewed circular trajectory. The results of these experiments show that the type-2 fuzzy controller provides better transient and steady-state response compared to a conventional fuzzy controller.
{"title":"Real-Time Implementation of an Interval Type-2 Fuzzy Logic Controller for the Trajectory Tracking of an UAV","authors":"K. O. Canbek, Y. Oniz","doi":"10.1109/ISMSIT52890.2021.9604539","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604539","url":null,"abstract":"In this study, a Type-2 Fuzzy Logic Controller that directly generates the necessary control signals is used to achieve the trajectory tracking control of an unmanned aerial vehicle (UAV). The proposed control scheme takes the quadcopter’s divergence from the given trajectory and its derivative as inputs. A type-1 fuzzy logic controller is also designed and implemented on the quadcopter to assess the performance of the type-2 controller. Real-time flight experiments are conducted for a skewed circular trajectory. The results of these experiments show that the type-2 fuzzy controller provides better transient and steady-state response compared to a conventional fuzzy controller.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"36 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120844487","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-21DOI: 10.1109/ISMSIT52890.2021.9604645
Alper Tunga Akın, Ç. Cömert
Augmented reality applications running on smartphones or tablets are becoming increasingly common. It is crucial to extract the physical structure of the scene perceived by the device camera in these applications. In such applications, employed in education, navigation and obstacle notification, the distance information between the device camera and the object must be derived and processed with sufficient accuracy and speed. In this study, the deep learning model, named "From Big To Small (BTS)", with superior performance metrics in depth extraction according to the literature reviews, was transformed into a web service and tested on an Android phone. Thus, a deep learning model with a high computational cost will be available on an Android device with average processing power. Test results were examined, and improvements were discussed.
在智能手机或平板电脑上运行的增强现实应用程序正变得越来越普遍。在这些应用中，提取设备摄像机感知到的场景的物理结构是至关重要的。在这些应用中，用于教育、导航和障碍物通知，设备相机与物体之间的距离信息必须以足够的精度和速度导出和处理。在本研究中，根据文献综述，我们将深度提取的性能指标从大到小(From Big To Small，简称BTS)深度学习模型转化为web服务，并在Android手机上进行测试。因此，一个计算成本高的深度学习模型将在处理能力一般的Android设备上可用。对试验结果进行了检验，并讨论了改进措施。
{"title":"Testing of a Deep Learning Model Providing Monocular Depth Estimation on Mobile Devices via Web Service","authors":"Alper Tunga Akın, Ç. Cömert","doi":"10.1109/ISMSIT52890.2021.9604645","DOIUrl":"https://doi.org/10.1109/ISMSIT52890.2021.9604645","url":null,"abstract":"Augmented reality applications running on smartphones or tablets are becoming increasingly common. It is crucial to extract the physical structure of the scene perceived by the device camera in these applications. In such applications, employed in education, navigation and obstacle notification, the distance information between the device camera and the object must be derived and processed with sufficient accuracy and speed. In this study, the deep learning model, named \"From Big To Small (BTS)\", with superior performance metrics in depth extraction according to the literature reviews, was transformed into a web service and tested on an Android phone. Thus, a deep learning model with a high computational cost will be available on an Android device with average processing power. Test results were examined, and improvements were discussed.","PeriodicalId":120997,"journal":{"name":"2021 5th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127043445","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：