Information and telecommunication sciences: international research journal

Постійне посилання на фонд

https://ela.kpi.ua/handle/123456789/11750
ISSN 2411-2976 (Online), ISSN 2312-4121 (Print) Періодичність: 2 рази на рік
Рік заснування: 2010
Тематика: теорія телекомунікацій та обробка сигналів; побудова сучасних та перспективних мереж; безпроводові технології та системи; керування в системах та мережах телекомунікацій; моделювання та оптимізація систем і мереж; програмні засоби та інформаційні ресурси телекомунікацій; кабельні та волоконно-оптичні системи; мікрохвильова техніка та терагерцові технології; історія телекомунікацій.
Попередня назва: Telecommunication Sciences (до 2013 року)
Офіційний сайт: http://infotelesc.kpi.ua/

Переглянути

Перегляд Information and telecommunication sciences: international research journal за Ключові слова "004.391"

Зараз показуємо 1 - 2 з 2
  • Ескіз
    ДокументВідкритий доступ
    Implementation of technology for improving the quality of segmentation of medical images by software adjustment of convolutional neural network hyperparameters
    (National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute", 2023) Prochukhan, Dmytro V.
    Background. The scientists have built effective convolutional neural networks in their research, but the issue of optimal setting of the hyperparameters of these neural networks remains insufficiently researched. Hyperparameters affect model selection. They have the greatest impact on the number and size of hidden layers. Effective selection of hyperparameters improves the speed and quality of the learning algorithm. It is also necessary to pay attention to the fact that the hyperparameters of the convolutional neural network are interconnected. That is why it is very difficult to manually select the effective values of hyperparameters, which will ensure the maximum efficiency of the convolutional neural network. It is necessary to automate the process of selecting hyperparameters, to implement a software mechanism for setting hyperparameters of a convolutional neural network. The author has successfully implemented the specified task. Objective. The purpose of the paper is to develop a technology for selecting hyperparameters of a convolutional neural network to improve the quality of segmentation of medical images. Methods. Selection of a convolutional neural network model that will enable effective segmentation of medical images, modification of the Keras Tuner library by developing an additional function, use of convolutional neural network optimization methods and hyperparameters, compilation of the constructed model and its settings, selection of the model with the best hyperparameters. Results. A comparative analysis of U-Net and FCN-32 convolutional neural networks was carried out. U-Net was selected as the tuning network due to its higher quality and accuracy of image segmentation. Modified the Keras Tuner library by developing an additional function for tuning hyperparameters. To optimize hyperparameters, the use of the Hyperband method is justified. The optimal number of epochs was selected - 20. In the process of setting hyperparameters, the best model with an accuracy index of 0.9665 was selected. The hyperparameter start_neurons is set to 80, the hyperparameter net_depth is 5, the activation function is Mish, the hyperparameter dropout is set to False, and the hyperparameter bn_after_act is set to True. Conclusions. The convolutional neural network U-Net, which is configured with the specified parameters, has a significant potential in solving the problems of segmentation of medical images. The prospect of further research is the use of a modified network for the diagnosis of symptoms of the coronavirus disease COVID-19, pneumonia, cancer and other complex medical diseases.
  • Ескіз
    ДокументВідкритий доступ
    Іmplementation of medical mask recognition technology in real time using a video camera
    (National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, 2022) Prochukhan, Dmytro; Korablev, Mykola
    Background. The coronavirus pandemic (COVID-19) has led to a major global health crisis due to the rapid spread of the virus. The World Health Organization has provided guidelines for protection against this disease. One of the effective recommendations is to wear a medical mask in public or crowded places. Involving human resources to monitor how these requirements are met is ineffective. It is necessary to automate the process of determining a properly dressed medical mask in real time using a video camera. The authors have developed an application that effectively copes with this task. Objective. The purpose of thew paper is the application development of a medical mask recognition in real time using a video camera, adapted for the use in modern Ukrainian reality with high accuracy and low system requirements. Methods. Analysis of existing analogues in the world, building a model of convolutional neural network, the architecture of which will detect and classify the image obtained from the camcorder in real time, create application architecture, develop a model in Python programming language, application testing. Results. A convolutional neural network of its own architecture has been created. The use of the Adam algorithm to optimize learning and use binary cross entropy as a cost function is substantiated. The method of face recognition using Haar features has been improved. High rate of convolutional neural network training was obtained: the training set accuracy - 97.46%, the test set accuracy - 97.23%, the cost function value at the training set - 2.37%, the cost function value at the test set - 2.57%. An application consisting of three modules has been developed: a machine learning module, an image processing module, a video camera activation module and a mask recognition device. Conclusions. The application effectively copes with the task of recognizing the presence of a medical mask in real time. The developed model in comparison with the analogues has a smaller size and simpler architecture without compromising accuracy or speed. The software has been successfully tested on various operating systems and hardware configurations. The application can be used in areas where there is a need to automate the process of determining the presence and accuracy of wearing a medical mask in real time.