Volume 32 #3

Contents

1. Fundamental principles of odor perception for higher fatty aids V. Ì. Ganshina and E. P. Zinkevich
2. Attractors for visual attention and visual scene analysis
3. Influence of the pre-exposure of visual objects on the discrimination of sounds in children with a delay in mental and language development
4. Neural network-based feature point descriptors for registration of optical and SAR images
5. Artificial training data generation for the task of character recognition of fields of russian passport
6. Methods of training data augmentation in the task of image classification
7. Neural network model of multisensory detector vehicle presence in the classification zone of charging points
8. Learning a metric feature space with siamese networks for disparity map computation
9. A method of object recognition based on formal properties specification applied to locating objects in aerial photographs

The paper considers the issue of event recognition by a fully connected artificial neural network. Input data present by time series, contains information about sensors condition. Using fully connected neural network, not recurrent, requires compression input time series. It is necessary to reduce count of input neurons. In the work is presented compression algorithm. This algorithm is analog run-length encoding algorithm with some changes. Detector, building by neural network was learning on real data from automatic vehicle classifier. The detector has high accuracy.

Key words: fully connected neural networks, time series, detection extended event

DOI: 10.1134/S0235009218030095

References:

• Bocharov D., Koptelov I., Kuznetsova E. Detektory proezdov na osnove tekhnicheskogo zreniya v avtomaticheskom klassifikatore transportnykh sredstv [Machine vision based vehicle pass detectors in automatic vehicle classifier]. Informatsionnye tekhnologii i sistemy [Information Technology and Systems]. 2015. P. 485–497 (in Russian).
• Burnaev E., Koptelov I., Novikov G., Khanipov T. Avtomatizirovannoe postroenie klassifikatorov na osnove rekurrentnykh neironnykh setei dlya detektsii proezdov transportnykh sredstv [Automatic Construction of a Recurrent Neural Network based Classifier for Vehicle Passage Detection]. Informatsionnye tekhnologii i sistemy [Information Technology and Systems]. 2016. P. 57–62 (in Russian).
• Grigoryev A., Gladilin S., Khanipov T., Koptelov I., Bocharov D., Matsnev D. Arkhitektura sistemy detektsii i klassifikatsii avtomobilei sredstvami tekhnicheskogo zreniya v estestvennykh usloviyakh [Architecture of a system for computer vision-based vehicle detection and classification under natural conditions]. Sensornye sistemy [Sensory systems]. 2017. V. 31(1). P. 72–84 (in Russian).
• Koptelov I., Grigoryev A., Khanipov T., Emelyanov S., Nikolaev D. Maket avtomaticheskogo klassifikatora transportnykh sredstv [Automatic vehicle classifier model]. Informatsionnye tekhnologii i sistemy [Information Technology and Systems]. 2014. P. 350–356 (in Russian).
• Malyugina O., Nikolaev D. Kriterii otsenki kachestva dlya potokovoi sistemy obnaruzheniya i klassifikatsii [Quality measurement criteria for detection and classification streaming system]. Informatsionnye tekhnologii i sistemy [Information Technology and Systems]. 2015. P. 414–427 (in Russian).
• Khanipov T., Koptelov I., Grigoryev A., Kuznetsova E., Nikolaev D. Vision-based industrial automatic vehicle classifier. Seventh International Conference on Machine Vision. 2015. V. 9445(944511). P. 1–5.
• Lin C.T., Lee C.S. G. Neural-network-based fuzzy logic control and decision system. IEEE Transactions on computers. 1991. V. 40(12). P. 1320–1336. 10.1109/12.106218.
• Robinson A.H., Cherry C. Results of a prototype television bandwidth compression scheme. Proceedings of the IEEE. 1967. 55(3). P. 356–364. 10.1109/PROC. 1967.5493.