Simultaneous localisation and mapping (SLAM) is an essential component of robotic systems. In this work we perform a feasibility study RGB-D SLAM for the task indoor robot navigation. Recent visual methods, e.g. ORBSLAM2 [9], demonstrate really impressive accuracy, but experiments in papers are usually conducted on just few sequences, that makes it difficult to reason about robustness methods. Another problem all available datasets contain trajectories with very complex camera motions....

Purpose The present paper aims to explore how measure trust as a receptivity force in an intra-organisational knowledge-sharing network with the help of self-developed algorithms modelling percolations. Design/methodology/approach In this paper, completely new methodology is applied by using sample study international company’s financial centre example. Computer software has been developed simulate and calculate percolation thresholds combining its characteristics, thereby revealing what...

This article describes the solution of two problems. First, based on fractional diffusion equation, a boundary problem with arbitrary values derivative indicators was formulated and solved, describing more general cases than existing solutions. Secondly, from consideration probability schemes transitions between states process, which can be observed in complex systems, fractional-differential equation telegraph type multiples is obtained (in time: β, 2β, 3β, … state: α, 2α, 3α, …) using...

The urgency of conducting research in the field quantum informatics is grounded. Promising areas are highlighted. For foreign and Russian publications materials, a review main scientific results that characterize current state computer science made. It noted knowledge funds invested most intensively development architecture its elements. Despite fact today there no information on creation physical implementation comparable functionality to classical digital computer, algorithms one popular...

We present a novel dataset for training and benchmarking semantic SLAM methods. The consists of 200 long sequences, each one containing 3000-5000 data frames. generate the sequences using realistic home layouts. For that we sample trajectories simulate motions simple robot, then render frames along trajectories. Each frame contains a) RGB images generated physically-based rendering, b) simulated depth measurements, c) IMU readings d) ground truth occupancy grid house. Our serves wider range...

This paper proposes a novel method of forecasting changes in social network systems based on applying percolation theory and stochastic dynamics approach. New results from computer modeling the influence density its threshold are discussed. Percolation thresholds calculated for various densities can be used models describing system's transition one state to another. For example, with help developed models, it is possible forecast change time share people certain opinions views depending...

To describe the structure of news information domain, we introduce concept a director, that is conditional axis, position which defined by averaging directions vectors set location all clusters centroids. characterise dynamics reaching event, concepts decreasing and increasing trends domain varying. The analysis self-similarity in behaviour directors on basis proposed model can allow detecting periodicity approaching or moving away from point predicted event.

Abstract The paper shows that to study the reliability and fault tolerance of air transportation networks, methods percolation theory can be used, in which any aviation transport structure represented as a random non-planar, incompletely connected graph (nodes are airports, arcs airlines). In percolation, one consider solution problems finding shares blocked nodes connections for networks with various regular structures, they decompose into unconnected areas. share (in node problem) or...

One of the problems forecasting events in news feeds, is development models which allow for work with semi structured information space text documents. This article describes a model based on use stochastic dynamics changes structure non-stationary time series clusters (states space) basis diffusion approximation. Forecasting feed their description, vectorization, and finding cosine value angle between given vector centroids various semantic clusters. Changes over such angles above can be...

The paper expands on the elaborated model of forecasting news events basing stochastic dynamics clusters variance and memory-based realisation in an information domain with self-organizing poorly structured data. provided analysis overcoming threshold shows transition probability appearing right after process changing dimensions structure begins. allowances for second time derivative differential enable accounting memory previous states system possibility to describe system's considerable...

We consider the issues of developing innovative theoretical approaches to describe states and dynamics public mood opinions in social networks, further on envision practical application models obtained with help sociological research forecast people's behaviour management various communities society. The paper amply demonstrates workability percolation for such purposes. For example, they can define threshold values negative attitudes society study how clusters into particular groups united...

This paper explores the social dynamics of processes in complex systems involving humans by focusing on user activity online media outlets. The R/S analysis showed that time series under consideration are fractal and anti-persistent (they have a short-term memory Hurst exponent significantly less than 0.5). Following statistical processing, observed data there is small amount asymmetry distribution change amplitudes news comments; amplitude almost symmetrical, but heavy tail as probability...

Analyses of the processes information transfer within network structures shows that conductivity and percolation threshold depend not only on its density (average number links per node), but also spatial symmetry groups topological dimension. The results presented in this paper regarding simulation show that, for regular random 2D 3D networks, an increase (density) node reduces their value. At same density, value is less than whatever structure may be. Regardless type networks symmetry,...

It is required to assess the hydraulic unit current state, if it serviceable or unserviceable, as per its vibration monitoring data. This task of technical diagnostics, i.e. determine system state under conditions limited information, obtained during operation. Machine learning can be utilized solve this problem. During investigation cross-validation was applied for different variants original samples drilling-down, plotting classification models, and selection significant exponents. Taking...

For a description of the information space it is introduced vector representation constituent text documents that are bound by events described in timeline.The predicted event also represented obtained on base its description.The mean value projections direction at different time points considered as set system states.It entered change values states.To describe transitions between states used probabilistic approach and difference transition scheme.This makes possible to get dependence for...

The paper considers the solution of a fractional differential boundary value problem, that is, diffusion-type equation with arbitrary values derivative orders on an infinite axis. difference between obtained results and other authors’ ones is these involve orders. solutions described in literature, as rule, are considered case when time β lies range: 0 < ≤ 1, state α (the variable describing process) 1 2. presented article allows us to consider any ranges for β, if inequality β/α 0.865...

In recent decades, the role of socio-economic conditions for development internationalization higher education has grown rapidly.In article, authors analyzed definition concept "internationalization education", identified prerequisites and factors that influence process in field education.Based on a comparative analysis features West, Russia China, it is concluded studied countries there are differences education.

Learning-based visual odometry and SLAM methods demonstrate a steady improvement over past years. However, collecting ground truth poses to train these is difficult expensive. This could be resolved by training in an unsupervised mode, but there still large gap between performance of supervised methods. In this work, we focus on generating synthetic data for deep learning-based that take optical flow as input. We produce form corresponds arbitrary camera movement real frame virtual frame....