Václav Snáel

Map Similarity Testing Using Matrix Decomposition

The similarity of two maps can be most easily compared visually. In this case, the degree of similarity is very subjective. It is therefore necessary to find an objective method for measuring similarity. This paper presents a method based on Singular Value Decomposition (SVD).

Muzzle-Based Cattle Identification Using Speed up Robust Feature Approach

Starting from the last century, animals identification became important for several purposes, e.g. tracking, controlling livestock transaction, and illness control. Invasive and traditional ways used to achieve such animal identification in farms or laboratories. To avoid such invasiveness and to get more accurate identification results, biometric identification methods have appeared. This paper presents an invariant biometric-based identification system to identify cattle based on their muzzle print images. This system makes use of Speeded Up Robust Feature (SURF) features extraction technique along with with minimum distance and Support Vector Machine (SVM) classifiers. The proposed system targets to get best accuracy using minimum number of SURF interest points, which minimizes the time needed for the system to complete an accurate identification. It also compares between the accuracy gained from SURF features through different classifiers. The experiments run 217 muzzle print images and the experimental results showed that our proposed approach achieved an excellent identification rate compared with other previous works.

Human Thermal Face Recognition Based on Random Linear Oracle (RLO) Ensembles

This paper proposes a human thermal face recognitionapproach with two variants based on Random linearOracle (RLO) ensembles. For the two approaches, the Segmentation-based Fractal Texture Analysis (SFTA) algorithmwas used for extracting features and the RLO ensembleclassifier was used for recognizing the face from its thermalimage. For the dimensionality reduction, one variant (SFTALDA-RLO) was used the technique of Linear DiscriminantAnalysis (LDA) while the other variant (SFTA-PCA-RLO) wasused the Principal Component Analysis (PCA). The classifiersmodel was built using the RLO classifier during the trainingphase and in the testing phase then this model was usedto identify the unknown sample images. The two variantswere evaluated using the Terravic Facial IR Database and theexperimental results showed that the two variants achieved agood recognition rate at 94.12% which is better than related work.

Visualization of Large Graphs Using GPU Computing

Graphs may be used to visualize relationships between objects. Relations are represented by edges and objects are called nodes. When graph is drawn, one can easily see and understand the basic structure of data. Many different applications can be found in social network analysis, computer networks, scientific literature analysis, etc. However drawing large graphs (thousands or a millions of nodes), is still challenging problem. There exist many different algorithms for drawing graphs. Each algorithm has specific behavior and different applications and limits. Some algorithms are focused on quality while others are more suitable for large graphs. This paper aims to speed up the computation using GPU, so larger graphs can be visualized in acceptable time, or visualization can be done even in real-time.

Modeling Permutations for Genetic Algorithms

Combinatorial optimization problems form a class of appealing theoretical and practical problems attractive for their complexity and known hardness. They are often NP-hard and as such not solvable by exact methods. Combinatorial optimization problems are subject to numerous heuristic and metaheuristic algorithms, including genetic algorithms. In this paper, we present two new permutation encodings for genetic algorithms and experimentally evaluate the influence of the encodings on the performance and result of genetic algorithm on two synthetic and real-world optimization problems.

Reducing Social Network Dimensions Using Matrix Factorization Methods

Since the availability of social networks data and the range of these data have significantly grown in recent years, new aspects have to be considered. In this paper we address computational complexity of social networks analysis and clarity of their visualization. Our approach uses combination of Formal Concept Analysis and well-known matrix factorization methods. The goal is to reduce the dimension of social network data and to measure the amount of information which is lost during the reduction.

Scheduling Independent Tasks on Heterogeneous Distributed Environments by Differential Evolution

Scheduling is one of the core steps to efficiently exploit the capabilities of heterogeneous distributed computing systems and it is also an appealing NP-complete problem. There is a number of heuristic and meta-heuristic algorithms that were tailored to deal with scheduling of independent jobs. In this paper we investigate the efficiency of differential evolution on the scheduling problem.

Neural Networks for Emotion Recognition Based on Eye Tracking Data

We present an approach for emotion recognition using information of the pupil. In last years, the pupil variables have been used as an assessment of emotional arousal. In this article, we generate signals of pupil size and gaze position monitored during image viewing. The emotions are provoked by visual stimuli of colored images. Those images were taken from the International Affective Picture System which has been the reference for objective emotional assessment based on visual stimuli. For recognising the emotions we use the evolution of the eye tracking data during a window of time. The learning dataset is composed by the evolution of the pupil size and the gaze position, and labels associated to the emotional states. We study two kinds of learning tools based on Neural Networks. We obtain promising empirical results that show the potential of using temporal learning tools for emotion recognition.

Local Dependency in Networks

Many real world data or process eave a network structure and can usefully be represented as graphs. Network analysis focuses on the relations among the nodes exploring the properts hies of network. We introduce a method for measuring the dependency between the nodes of a network, that is based on a structure in the local surroundings of the node. The approach extracts relations between the networks nodes and from either unweighted or already weighted network we get a weighted network where the assigned edge weights reflect the dependency between the nodes. Additionally, from dependency between the nodes, we derive a novel degree centrality measure which provides an interesting view on the importance of the node in a network.

Randomness and Chaos in Genetic Algorithms and Differential Evolution

Evolutionary methods and stochastic algorithms in general rely heavily on streams of (pseudo-)random numbers generated in course of their execution. The pseudo-random numbers are utilized for in-silico emulation of probability-driven natural processes such as modification of genetic information (mutation, crossover), partner selection, and survival of the fittest (selection, migration). Deterministic chaos is a very well known mathematical concept that can be used to generate sequences of real numbers within selected interval. In the past, it has been used as a basis for various pseudo-random number generators with interesting properties. This work provides an empirical comparison of the performance of genetic algorithms and differential evolution using different pseudo-random number generators and chaotic systems as sources of stochasticity.