2687-0517 11 2 2018

RAR RUS 7-18 Peter the Great St.Petersburg Polytechnic University Makarov Aleksei lyohamakarov@yandex.ru

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Peter the Great St.Petersburg Polytechnic University Bolsunovskaya Marina bolsun_hht@mail.ru

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Comparative analysis of methods for detecting special points in images at different levels of illumination In this article, we have carried out a comparative analysis of methods for detecting special points in an image, which is part of the study on developing an around view system for large vehicles. Because nighttime is especially dangerous for driving and most difficult for stitching images, this article focused on the possibility of detecting special points and stitching in low illumination. We made a comparative analysis of the methods for detecting special points in images, developed a technique and conducted experiments to find special points in images using such methods as SURF, MSER, BRISK, Harris, FAST and MinEigen. During the study, we have performed a search for identical special points for a pair of images, an analysis of their number and stitching of images by different methods at different levels of illumination. The results of the experiments are given in graphs and tables. 10.18721/JCSTCS.11201 004.932 detecting special points stitching images low level of illumination comparison technique SURF https://infocom.spbstu.ru/article/2018.57.1/ 2018_2_01.pdf

RAR RUS 19-31 Pyattaev Alexander RUDN University Andreev Sergey serge.andreev@gmail.com

6, Miklukho-Maklaya str., Moscow, 117198, Russia

D-5155-2014 6507253900 0000-0003-3976-2971 Higher School of Economics Yevgeni A. ykoucheryavy@hse.ru

Korkeakoulunkatu 10, FI-33720 Tampere Finland

Design and evaluation of a system of direct connections D2D with LTE cellular assistance Due to the growing popularity of mobile user applications, cellular network operators are increasingly interested in offloading their scarce licensed spectrum resources. To this aim, using unlicensed-band radio technologies is attractive owing to its reduced costs. This work designs a system of direct D2D (device-to-device) connections to offload mobile traffic that is served by LTE (long-term evolution) cellular networks. In particular, a network assistance protocol over LTE architecture is proposed for D2D connections, and its system-level performance is evaluated to quantify the resulting traffic-steering efficiency. In addition, practical operation of D2D communication in a test LTE network is described, which allows making substantiated conclusions about the actual feasibility of the proposed solution in real-life conditions. 10.18721/JCSTCS.11202 004.77 mobile traffic LTE cellular communication direct connections with network assistance WiFi Direct radio technology design and evaluation https://infocom.spbstu.ru/article/2018.57.2/ 2018_2_02.pdf

RAR RUS 35-46 Saint Petersburg State University Moiseenko Evgenii e.moiseenko@2012.spbu.ru

7-9 Universitetskaya Emb., St Petersburg 199034, Russia

Saint Petersburg State University Podkopaev Anton Relational programming with memoization and negation The relational paradigm allows to express programs as relations. Unlike functions, the relations do not distinguish input and output parameters, thus a single relational program can be used to solve several related problems. Relational interpreters are of particular interest. These interpreters can execute a program, check that the program satisfies a set of constraints or generate a program that has specified properties. In order to take advantages of the relational interpreter, the developer needs to define the semantics of the programming language as a relation. In this work, we present an implementation of two useful extensions of relational programming: tabling and constructive negation. Tabling helps to traverse the state space of the interpreter efficiently. Constructive negation allows to check that some state of the interpreter is unreachable. We show how this extensioncan be used on an example of a relational interpreter for a concurrent imperative programming language. 10.18721/JCSTCS.11203 004.434 relational programming declarative programming constraint logic programming tabling constructive negation https://infocom.spbstu.ru/article/2018.57.3/ 2018_2_03.pdf

RAR RUS 47-53 Norilsk State Industrial Institute Sulyaev Ilgiz ilgizfinland@mail.ru “Norilsk Nickel” Sedov Dmitriy dvsedov95@mail.ru

Norilsk

Modeling of a low-pressure air separator as a control object The article deals with the task of developing a mathematical model of an air separator as the control object. We have carried out systematic analyses of the object and built a conceptual model with the principle of revelation consistency according to vagueness. The results of active and passive identification are presented. The analysis of the experiment results is carried out. The causes of divergences, which are determined mainly because of structural peculiarities, are established. 10.18721/JCSTCS.11204 681.5 air-separation oxygen identification mathematical model systematic analysis https://infocom.spbstu.ru/article/2018.57.4/ 2018_2_04.pdf

RAR RUS 55-63 Peter the Great St.Petersburg Polytechnic University Tian Zhaolin peter0431peter@gmail.com

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Peter the Great St.Petersburg Polytechnic University Zhang Weiwei soszhang@outlook.com

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Sequence-to-Sequence Based English-Chinese Translation Model In recent years, with the continuous improvement of theory in artificial intelligence, artificial neural networks has become novel tools for machine translation. Compared with traditional Statistical Machine Translation (SMT), neural network based Neural Machine Translation (NMT) transcends SMT in many aspects such as translation accuracy, long distance reordering, syntax, tolerance to noisy data et al. In 2014, with the emergence of sequence-to-sequence (seq2seq) models and attentional mechanisms introduced into the model, NMT was further refined and its performance was getting better and better. This article uses the current popular sequence-to-sequence model to construct a neural machine translation model from English to Chinese. In addition, this paper uses Long-Short Term Memory (LSTM) to replace the traditional RNN in order to solve the problem of gradient disappearance and gradient explosion that it faces in long-distance dependence. The attention mechanism has also been introduced into this article. It allows neural networks to pay more attention to the relevant parts of the input sequences and less to the unrelated parts when performing prediction tasks. In the experimental part, this article uses TensorFlow to build the NMT model described in the article. 10.18721/JCSTCS.11205 004 NMT seq2seq LSTM attention mechanism encoder-decoder TensorFlow https://infocom.spbstu.ru/article/2018.57.5/

RAR RUS 64-71 Peter the Great St.Petersburg Polytechnic University Sichkar Valentin valentyn.s2014@yandex.ru

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Comparison analysis of knowledge-based systems for navigation of mobile robot and collision avoidance with obstacles in unknown environment Developing systems for intelligent navigation is one of the major problems in world of modern robotics. This problem is particularly urgent when the environment is unknown. It means that a mobile robot meeting unpredictable obstacles on its way and has to react according to the current situation fast and in real time. That is why developing such a system is always a big challenge. This paper studies different techniques for storing and using the knowledge in order to avoid collisions with obstacles. Most attention is paid for developing two types of Knowledge Bases to help the mobile robot to avoid possible collisions and continue its way. A comparison analysis is provided for these two different types of Knowledge Bases. The advantages and disadvantages were analyzed and described. 10.18721/JCSTCS.11206 004.81 mobile robot intelligent navigation obstacle avoidance symbolic knowledge base neural network knowledge base https://infocom.spbstu.ru/article/2018.57.6/