Recognizing atypical on-the-go situations using a convolutional neural network

Authors

  • O.K. Kolesnytskiy Vinnytsia National Technical University
  • S.V. Kykynin Spotlight Media Labs, Sunnyvale, California
  • M.Yu. Derevyanko Vinnytsia National Technical University
  • A.A. Prepodobnyy Mendesh Da Maya Vinnytsia National Technical University

DOI:

https://doi.org/10.31649/1681-7893-2019-38-2-38-44

Keywords:

image processing, atypical situations, traffic accidents, convolutional neural network

Abstract

The information technology of atypical situations recognition on the road is offered and its software implementation is investigated. YOLO convolutional neural network was used to detect and track objects on the road in real time. To identify atypical situations, we used an analysis of changes in the motion characteristics of detected objects.

Author Biographies

O.K. Kolesnytskiy, Vinnytsia National Technical University

доцент, канд. техн. наук, доцент кафедри комп’ютерних наук

S.V. Kykynin, Spotlight Media Labs, Sunnyvale, California

Principal Full Stack Software Engineer

M.Yu. Derevyanko, Vinnytsia National Technical University

магістрант кафедри комп’ютерних наук

A.A. Prepodobnyy Mendesh Da Maya, Vinnytsia National Technical University

магістрант кафедри комп’ютерних наук

References

Road accident statistics in Ukraine [Electronic resource]. - Access mode: http://patrol.police.gov.ua/statystyka/

Safe City: Over 400 surveillance cameras have been installed in Vinnitsa [Electronic resource]. - Accessmode: https://www.myvin.com.ua/en/news/events/53372.html

Frank Millstein Convolutional Neural Networks in Python: Beginner's Guide to Convolutional NeuralNetworks in Python. CreateSpace Independent Publishing Platform, 2018. 120 p.

Sight Machine. Powering digital Manufacturing. - Access mode: https://sightmachine.com/.

Deep Neural Networks for Image Recognition and Classification Problems [Electronic resource]. - Accessmode: http://itcm.comp-sc.if.ua/2017/Sineglazov.pdf.

Convolutional neural network - a simple explanation of CNN and its application [Electronic resource]. -Access mode: https://evergreens.com/en/articles/cnn.html

Artificial Intelligence (AI) Vs. Machine Learning Vs. Deep Learning [Online resource]. - Access mode:https://skymind.ai/wiki/ai-vs-machine-learning-vs-deep-learning.

Darknet: Open Source Neural Networks and YOLO: Real-Time Object Detection. - Access mode:https://pjreddie.com/darknet/yolo.

Shcherbakova G.Yu. Probability theory lecture notes. Odessa: Science and Technology, 2005. 68 p.

VF Bardachenko, OK Kolesnitsky, SA Vasiletsky. Prospects for the use of pulsed neural networks withtimer representation of information for dynamic pattern recognition // USiM.-2003-№6.- P. 73-82.

Kolesnitsky OK Analytical review of hardware realizations of spike neural networks / OK Kolesnitsky //Mathematical Machines and Systems. - 2015. - №1, P.3-19. ISSN 1028-9763 [Electronic resource]. AccessMode - http://www.immsp.kiev.ua/publications/articles/2015/2015_1/01_2015_Kolesnytskyy.pdf

Neurocomputer architecture based on spiking neural network and its optoelectronic implementation / OlehK. Kolesnytskyj; Vladislav V. Kutsman; Krzysztof Skorupski; Mukaddas Arshidinova, Proc. SPIE 11176,Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments2019, 1117609 (6 November 2019); doi: 10.1117 / 12.2536607

Downloads

Abstract views: 401

Published

2020-03-12

How to Cite

[1]
O. Kolesnytskiy, S. Kykynin, M. Derevyanko, and A. Prepodobnyy Mendesh Da Maya, “Recognizing atypical on-the-go situations using a convolutional neural network”, Опт-ел. інф-енерг. техн., vol. 38, no. 2, pp. 38–44, Mar. 2020.

Issue

Section

Systems Of Technical Vision And Artificial Intelligence, Image Processing And Pattern Recognition

Metrics

Downloads

Download data is not yet available.