Convolutional neural networks of the YOLO class in computer vision systems for mobile robotic complexes; Control and Communications (SIBCON-2019)
| Parent link: | Control and Communications (SIBCON-2019).— 2019.— [18739806, 5 p.] |
|---|---|
| Hlavní autor: | |
| Korporativní autor: | |
| Další autoři: | , |
| Shrnutí: | Title screen An important scientific direction is the development and study of computer vision systems (CVS) for mobile robotic complexes. Today, developers of CVS are most often using convolutional neural networks (CNN). For increasing the speed detection of objects on images in CVS, there has been a trend of using CNN, which are hardware-implemented on field-programmable gate array (FPGAs).This article shows that the perspective for hardware implementation on the FPGA is the tiny-YOLO CNN from the YOLO class. For reduce required FPGA computing resources in this CNN, was proposed to use Inception-ResNet modules. We was found that with high detection accuracy of objects in images with minimum resources requirements provide by the tiny-YOLO-Inception-ResNet2 network architecture. It is obtained from replacing the fifth tiny-YOLO convolutional layer of the tiny-YOLO CNN with two sequential processing Inception-ResNet modules. Also results of the study of the detection accuracy of objects using the CNN for this architecture with the lack of resource-intensive operations: batch normalization and bias from calculations were given. These studies were performed for different formats of representation numbers in the FPGA. Режим доступа: по договору с организацией-держателем ресурса |
| Jazyk: | angličtina |
| Vydáno: |
2019
|
| Témata: | |
| On-line přístup: | https://doi.org/10.1109/SIBCON.2019.8729605 |
| Médium: | MixedMaterials Elektronický zdroj Kapitola |
| KOHA link: | https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=661724 |
MARC
| LEADER | 00000naa0a2200000 4500 | ||
|---|---|---|---|
| 001 | 661724 | ||
| 005 | 20250829143452.0 | ||
| 035 | |a (RuTPU)RU\TPU\network\32555 | ||
| 090 | |a 661724 | ||
| 100 | |a 20200210d2019 k y0engy50 ba | ||
| 101 | 0 | |a eng | |
| 135 | |a drcn ---uucaa | ||
| 181 | 0 | |a i | |
| 182 | 0 | |a b | |
| 200 | 1 | |a Convolutional neural networks of the YOLO class in computer vision systems for mobile robotic complexes |f I. V. Zoev, A. P. Beresnev, N. G. Markov | |
| 203 | |a Text |c electronic | ||
| 300 | |a Title screen | ||
| 320 | |a [References: 10 tit.] | ||
| 330 | |a An important scientific direction is the development and study of computer vision systems (CVS) for mobile robotic complexes. Today, developers of CVS are most often using convolutional neural networks (CNN). For increasing the speed detection of objects on images in CVS, there has been a trend of using CNN, which are hardware-implemented on field-programmable gate array (FPGAs).This article shows that the perspective for hardware implementation on the FPGA is the tiny-YOLO CNN from the YOLO class. For reduce required FPGA computing resources in this CNN, was proposed to use Inception-ResNet modules. We was found that with high detection accuracy of objects in images with minimum resources requirements provide by the tiny-YOLO-Inception-ResNet2 network architecture. It is obtained from replacing the fifth tiny-YOLO convolutional layer of the tiny-YOLO CNN with two sequential processing Inception-ResNet modules. Also results of the study of the detection accuracy of objects using the CNN for this architecture with the lack of resource-intensive operations: batch normalization and bias from calculations were given. These studies were performed for different formats of representation numbers in the FPGA. | ||
| 333 | |a Режим доступа: по договору с организацией-держателем ресурса | ||
| 463 | |t Control and Communications (SIBCON-2019) |o proceedings of International Siberian Conference, April 18-20, 2019, Russia, Tomsk |v [18739806, 5 p.] |d 2019 | ||
| 610 | 1 | |a электронный ресурс | |
| 610 | 1 | |a труды учёных ТПУ | |
| 610 | 1 | |a mobile robotic systems | |
| 610 | 1 | |a computer vision systems | |
| 610 | 1 | |a object detection on images | |
| 610 | 1 | |a convolutional neural networks | |
| 610 | 1 | |a fieldprogrammable gate array | |
| 610 | 1 | |a роботизированные системы | |
| 610 | 1 | |a компьютерное зрение | |
| 610 | 1 | |a нейронные сети | |
| 700 | 1 | |a Zoev |b I. V. |c Specialist in the field of informatics and computer technology |c Programmer of Tomsk Polytechnic University |f 1993- |g Ivan Vladimirovich |3 (RuTPU)RU\TPU\pers\38250 | |
| 701 | 1 | |a Beresnev |b A. P. |c специалист в области информатики и вычислительной техники |c программист Томского политехнического университета |f 1994- |g Aleksey Pavlovich |3 (RuTPU)RU\TPU\pers\40132 | |
| 701 | 1 | |a Markov |b N. G. |c Doctor of Engineering, Professor of TPU |c Russian specialist in Informatics and Computing |f 1950- |g Nikolai Grigorevich |3 (RuTPU)RU\TPU\pers\24811 |9 11026 | |
| 712 | 0 | 2 | |a Национальный исследовательский Томский политехнический университет |b Инженерная школа информационных технологий и робототехники |b Отделение информационных технологий |3 (RuTPU)RU\TPU\col\23515 |
| 801 | 2 | |a RU |b 63413507 |c 20200210 |g RCR | |
| 856 | 4 | 0 | |u https://doi.org/10.1109/SIBCON.2019.8729605 |
| 942 | |c CF | ||