<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.3 20210610//EN" "JATS-journalpublishing1-3.dtd">
<article article-type="research-article" dtd-version="1.3" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xml:lang="ru"><front><journal-meta><journal-id journal-id-type="publisher-id">bsuir</journal-id><journal-title-group><journal-title xml:lang="ru">Доклады БГУИР</journal-title><trans-title-group xml:lang="en"><trans-title>Doklady BGUIR</trans-title></trans-title-group></journal-title-group><issn pub-type="ppub">1729-7648</issn><issn pub-type="epub">2708-0382</issn><publisher><publisher-name>БГУИР</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="doi">10.35596/1729-7648-2025-23-2-101-108</article-id><article-id custom-type="elpub" pub-id-type="custom">bsuir-4117</article-id><article-categories><subj-group subj-group-type="heading"><subject>Research Article</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="ru"><subject>Статьи</subject></subj-group></article-categories><title-group><article-title>Исследование аппаратной реализации нейронной сети прямого распространения для распознавания рукописных цифр на базе FPGA</article-title><trans-title-group xml:lang="en"><trans-title>Investigation of Hardware Implementation of a Feedforward Neural Network for Handwritten Digit Recognition Based on FPGA</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Кривальцевич</surname><given-names>Е. А.</given-names></name><name name-style="western" xml:lang="en"><surname>Krivalсevich</surname><given-names>E. A.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Студент</p><p>Минск</p></bio><bio xml:lang="en"><p>Egor A. Krivalcevich, Student</p><p>Minsk</p></bio><xref ref-type="aff" rid="aff-1"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Вашкевич</surname><given-names>М. И.</given-names></name><name name-style="western" xml:lang="en"><surname>Vashkevich</surname><given-names>M. I.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Вашкевич Максим Иосифович, д-р техн. наук, проф. каф. электронных вычислительных средств</p><p>220013, Минск, ул. П. Бровки, 6</p></bio><bio xml:lang="en"><p>Maxim I. Vashkevich, Dr.   Sci.   (Tech.),   Professor at the Electronic Computing Facilities Department</p><p>220013, Minsk, P. Brovki St., 6 </p></bio><email xlink:type="simple">vashkevich@bsuir.by</email><xref ref-type="aff" rid="aff-1"/></contrib></contrib-group><aff-alternatives id="aff-1"><aff xml:lang="ru"><institution>Белорусский государственный университет информатики и радиоэлектроники</institution></aff><aff xml:lang="en"><institution>Belarusian State University of Informatics and Radioelectronics</institution></aff></aff-alternatives><pub-date pub-type="collection"><year>2025</year></pub-date><pub-date pub-type="epub"><day>29</day><month>04</month><year>2025</year></pub-date><volume>23</volume><issue>2</issue><fpage>101</fpage><lpage>108</lpage><permissions><copyright-statement>Copyright &amp;#x00A9; Кривальцевич Е.А., Вашкевич М.И., 2025</copyright-statement><copyright-year>2025</copyright-year><copyright-holder xml:lang="ru">Кривальцевич Е.А., Вашкевич М.И.</copyright-holder><copyright-holder xml:lang="en">Krivalсevich E.A., Vashkevich M.I.</copyright-holder><license xml:lang="ru" license-type="creative-commons-attribution" xlink:href="https://creativecommons.org/licenses/by/4.0/" xlink:type="simple"><license-p>Данная работа распространяется под лицензией Creative Commons Attribution 4.0.</license-p></license><license xml:lang="en" license-type="creative-commons-attribution" xlink:href="https://creativecommons.org/licenses/by/4.0/" xlink:type="simple"><license-p>This work is licensed under a Creative Commons Attribution 4.0 License.</license-p></license></permissions><self-uri xlink:href="https://doklady.bsuir.by/jour/article/view/4117">https://doklady.bsuir.by/jour/article/view/4117</self-uri><abstract><p>Разработана аппаратная реализация на базе программируемых логических интегральных схем (ПЛИС) типа Field Programmable Gate Array однослойной нейронной сети прямого распространения для распознавания рукописных цифр. Исследовано влияние разрядности коэффициентов сети на точность распознавания и на аппаратные затраты ПЛИС. Обучение нейронной сети выполнялось с помощью базы рукописных цифр MNIST. Прототип нейронной сети был реализован в виде IP-ядра на отладочной плате ZYBO Z7. Разработанный прототип использовался для выполнения экспериментов с различной разрядностью представления коэффициентов нейронной сети. Построены графики точности распознавания и количества аппаратных ресурсов ПЛИС в зависимости от разрядности представления коэффициентов нейронной сети. Выполнен анализ полученных в результате обучения нейронной сети коэффициентов с использованием разложения на битовые плоскости. Показано, что для представления коэффициентов нейронной сети достаточно 5 разрядов, поскольку они содержат основную, усвоенную сетью, информацию, обеспечивая экономное расходование ресурсов ПЛИС и высокую точность распознавания (92,4 %).</p></abstract><trans-abstract xml:lang="en"><p>A hardware implementation based on Field Programmable Gate Array (FPGA) of a single-layer feedforward neural network for handwritten digit recognition has been developed. The effect of the network coefficient bit depth on the recognition accuracy and FPGA hardware costs has been studied. The neural network was trained using the MNIST handwritten digit database. The neural network prototype was implemented as an IP core on the ZYBO Z7 debug board. The developed prototype was used to perform experiments with different bit depths of neural network coefficient representation. Graphs of recognition accuracy and the amount of FPGA hardware resources depending on the bit depth of neural network coefficient representation have been constructed. The coefficients obtained as a result of neural network training have been analyzed using decomposition into bit planes. It has been shown that 5 bits are sufficient to represent neural network coefficients, since they contain the main information learned by the network, ensuring economical use of FPGA resources and high recognition accuracy (92.4 %).</p></trans-abstract><kwd-group xml:lang="ru"><kwd>нейронная сеть</kwd><kwd>распознание рукописных цифр</kwd><kwd>полносвязный слой</kwd><kwd>MNIST</kwd><kwd>FPGA</kwd><kwd>битовые плоскости</kwd></kwd-group><kwd-group xml:lang="en"><kwd>neural network</kwd><kwd>handwritten digit recognition</kwd><kwd>fully connected layer</kwd><kwd>MNIST</kwd><kwd>FPGA</kwd><kwd>bit planes</kwd></kwd-group></article-meta></front><back><ref-list><title>References</title><ref id="cit1"><label>1</label><citation-alternatives><mixed-citation xml:lang="ru">Mittal, S. A Survey of FPGA-Based Accelerators for Convolutional Neural Networks / S. Mittal // Neural Computing and Applications. 2020. Vol. 32, No 4. P. 1109–1139.</mixed-citation><mixed-citation xml:lang="en">Mittal S. (2020) A Survey of FPGA-Based Accelerators for Convolutional Neural Networks. Neural Computing and Applications. 32 (4), 1109–1139.</mixed-citation></citation-alternatives></ref><ref id="cit2"><label>2</label><citation-alternatives><mixed-citation xml:lang="ru">Ahmad, A. FFConv: An FPGA-Based Accelerator for Fast Convolution Layers in Convolutional Neural Networks / А. Ahmad, M. A. Pasha // ACM Transactions on Embedded Computing Systems (TECS). 2020. Vol. 19, Iss. 2. P. 1–24.</mixed-citation><mixed-citation xml:lang="en">Ahmad A., Pasha M. A. (2020) FFConv: An FPGA-Based Accelerator for Fast Convolution Layers in Convolutional Neural Networks. ACM Transactions on Embedded Computing Systems (TECS). 19 (2), 1–24.</mixed-citation></citation-alternatives></ref><ref id="cit3"><label>3</label><citation-alternatives><mixed-citation xml:lang="ru">FPGA Implementation of Hand-Written Number Recognition Based on CNN / D. Giardino [et al.] // International Journal on Advanced Science, Engineering and Information Technology. 2019. Vol. 9, No 1. P. 167–171.</mixed-citation><mixed-citation xml:lang="en">Giardino D., Matta M., Silvestri F., Spano S., Trobiani V. (2019) FPGA Implementation of Hand- Written Number Recognition Based on CNN. International Journal on Advanced Science, Engineering and Information Technology. 9 (1), 167–171.</mixed-citation></citation-alternatives></ref><ref id="cit4"><label>4</label><citation-alternatives><mixed-citation xml:lang="ru">Common Visual Data Foundation [Electronic Resource]. Mode of access: https://github.com/cvdfoundation/mnist.</mixed-citation><mixed-citation xml:lang="en">Common Visual Data Foundation. Available: https:// https://github.com/cvdfoundation/mnist.</mixed-citation></citation-alternatives></ref><ref id="cit5"><label>5</label><citation-alternatives><mixed-citation xml:lang="ru">Николенко, С. Глубокое обучение. Погружение в мир нейронных сетей / С. Николенко, А. Кадурин, Е. Архангельская. СПб.: Питер, 2019.</mixed-citation><mixed-citation xml:lang="en">Nikolenko S. I., Kadurin A. A., Arkhangelskaya E. V. (2019) Deep Learning. A Dive Into the World of Networks. Saint Petersburg, Peter Publishing House (in Russian).</mixed-citation></citation-alternatives></ref><ref id="cit6"><label>6</label><citation-alternatives><mixed-citation xml:lang="ru">Samaragh, M. Customizing Neural Networks for Efficient FPGA Implementation / М. Samaragh, M. Ghasemzadeh, F. Koushanfar // IEEE Symposium on Field-Programmable Custom Computing Machines. USA: California, 2017. P. 85–92.</mixed-citation><mixed-citation xml:lang="en">Samaragh M., Ghasemzadeh M., Koushanfar F. (2017) Customizing Neural Networks for Efficient FPGA Implementation. IEEE Symposium on Field-Programmable Custom Computing Machines. USA, California. 85–92.</mixed-citation></citation-alternatives></ref></ref-list><fn-group><fn fn-type="conflict"><p>The authors declare that there are no conflicts of interest present.</p></fn></fn-group></back></article>
