Skip to main content
main-content
Top

About this book

The LNCS journal Transactions on Computational Science reflects recent developments in the field of Computational Science, conceiving the field not as a mere ancillary science but rather as an innovative approach supporting many other scientific disciplines. The journal focuses on original high-quality research in the realm of computational science in parallel and distributed environments, encompassing the facilitating theoretical foundations and the applications of large-scale computations and massive data processing. It addresses researchers and practitioners in areas ranging from aerospace to biochemistry, from electronics to geosciences, from mathematics to software architecture, presenting verifiable computational methods, findings, and solutions, and enabling industrial users to apply techniques of leading-edge, large-scale, high performance computational methods.

This, the 36th issue of the Transactions on Computational Science, is devoted to the area of Cyberworlds and Cybersecurity. The first four papers constitute extended versions of selected papers presented at the 2018 International Conference on Cyberworlds, CW 2018. A further two papers were accepted following an open Call for Papers and cover the areas of fast 3D segmentation using geometric surface features and nature-inspired optimization for face recognition.

Table of Contents

Frontmatter

Orion: A Generic Model and Tool for Data Mining

Abstract
This paper focuses on the design of autonomous behaviors based on humans behaviors observation. In this context, the contribution of the Orion model is to gather and to take advantage of two approaches: data mining techniques (to extract knowledge from the human) and behavior models (to control the autonomous behaviors). In this paper, the Orion model is described by UML diagrams. More than a model, Orion is an operational tool allowing to represent, transform, visualize and predict data; it also integrates operational standard behavioral models. Orion is illustrated to control a bot in the game Unreal Tournament. Thanks to Orion, we can collect data of low level behaviors through three scenarios performed by human players: movement, long range aiming and close combat. We can easily transform the data and use some data mining techniques to learn behaviors from human players observation. Orion allows us to build a complete behavior using an extension of a Behavior Tree integrating ad hoc features in order to manage aspects of behavior that we have not been able to learn automatically.
Cédric Buche, Cindy Even, Julien Soler

Environment Estimation for Glossy Reflections in Mixed Reality Applications Using a Neural Network

Abstract
Environment textures are used for the illumination of virtual objects within a virtual scene. Using these textures is crucial for high-quality lighting and reflection. In the case of an augmented reality context, the lighting is very important to seamlessly embed a virtual object within the real world scene. To ensure this, the lighting of the environment has to be captured according to the current light information. In this paper, we present a novel approach by stitching the current camera information onto a cube map. This cube map is enhanced in every single frame and is fed into a neural network to estimate missing parts. Finally, the output of the neural network and the currently stitched information is fused to make even mirror-like reflections possible on mobile devices. We provide an image stream stitching approach combined with a neural network to create plausible and high-quality environment textures that may be used for image-based lighting within mixed reality environments.
Tobias Schwandt, Christian Kunert, Wolfgang Broll

Distance Measurements of CAD Models in Boundary Representation

Abstract
The need to analyze and visualize distances between objects arises in many use cases. Although the problem to calculate the distance between two polygonal objects may sound simple, real-world scenarios with large models will always be challenging, but optimization techniques – such as space partitioning – can reduce the complexity of the average case significantly.
Our contribution to this problem is a publicly available benchmark to compare distance calculation algorithms. To illustrate the usage, we investigated and evaluated a grid-based distance measurement algorithm.
Ulrich Krispel, Dieter W. Fellner, Torsten Ullrich

An Immersive Virtual Environment for Visualization of Complex and/or Infinitely Distant Territory

Abstract
With the advent of the high-performance graphics and networking technologies that enable us to create virtual worlds networked via the Internet, various virtual environments have been developed to support mathematics education at around the beginning of the 21st century. In the environments that have been inherently two- or three-dimensional Euclidean, students have discovered and experienced mathematical concepts and processes in almost the same ways that they can do in real life. Although elementary mathematics, for instance, calculus and linear algebra, plays an essential role in areas of understanding and knowledge to solve real-world problems, there are traditionally three general areas in pure mathematics for advanced problem-solving techniques: algebra, analysis, and geometry. So, using Virtual Reality (VR) as a general and advanced tool for mathematics education to support students not only in the primary and secondary, also in higher education, the virtual environment ideally provides a wide variety of mathematical domains as possible. We present an immersive virtual environment that allows the user to set environmental limits beyond three-dimensional Euclidean space. More specifically, by setting the limits to n-dimensional complex projective space, an element of both complex and infinitely distant domain can be naturally visualized as a recognizable form in the Euclidean 3-space. The problem here is that the higher the level of mathematics, the more the visualization method tends to become abstract that only experts with advanced degrees can fathom. We also show how our figurative approach is essential for bridging the gap between elementary and more advanced mathematical visualizations.
Atsushi Miyazawa, Masanori Nakayama, Issei Fujishiro

Fast 3D Scene Segmentation and Partial Object Retrieval Using Local Geometric Surface Features

Abstract
Robotic vision and in particular 3D understanding has attracted intense research efforts the last few years due to its wide range of applications, such as robot-human interaction, augmented and virtual reality etc, and the introduction of low-cost 3D sensing devices. In this paper we explore one of the most popular problems encountered in 3D perception applications, namely the segmentation of a 3D scene and the retrieval of similar objects from a model database. We use a geometric approach for both the segmentation and the retrieval modules that enables us to develop a fast, low-memory footprint system without the use of large-scale annotated datasets. The system is based on the fast computation of surface normals and the encoding power of local geometric features. Our experiments demonstrate that such a complete 3D understanding framework is possible and advantages over other approaches as well as weaknesses are discussed.
Dimitrios Dimou, Konstantinos Moustakas

Hybrid Nature-Inspired Optimization Techniques in Face Recognition

Abstract
Nature has been a very effective source to develop various Nature Inspired Optimisation algorithms and this has developed into an active area of research. The focus of this paper is to develop a Hybrid Nature-inspired Optimisation Technique and study its application in Face Recognition Problem. Two different hybrid algorithms are proposed in this paper. First proposed algorithm is a hybrid of Gravitational Search Algorithm (GSA) and Big Bang-Big Crunch (BBBC). The other algorithm is an improvement of the first algorithm, which incorporates Stochastic Diffusion Search (SDS) algorithm along with Gravitational Search Algorithm (GSA) and Big Bang-Big Crunch (BB-BC). The hybrid is an enhancement of a single algorithm which when incorporated with similar other algorithms performs better in situations where single algorithms fail to perform well. The algorithm is used to optimize the Eigen vectors generated from Principal Component Analysis. The optimized Eigen faces supplied to SVM classifier provides better face recognition capabilities compared to the traditional PCA vectors. Testing on the face recognition problem, the algorithm showed 95% accuracy in the ORL dataset and better optimization capability on functions like Griewank-rosenbrock, Schaffer F7 in comparison to standard algorithms like Rosenbrock, GA and DASA during the Benchmark Testing.
Lavika Goel, Abhilash Neog, Ashish Aman, Arshveer Kaur

Backmatter

Additional information

Premium Partner

    Image Credits