|
|
Learnable Evolution Model for Optimization (LEM)
Grunt, Pavel ; Vašíček, Zdeněk (referee) ; Schwarz, Josef (advisor)
My thesis is dealing with the Learnable Evolution Model (LEM), a new evolutionary method of optimization, which employs a classification algorithm. The optimization process is guided by a characteristics of differences between groups of high and low performance solutions in the population. In this thesis I introduce new variants of LEM using classification algorithm AdaBoost or SVM. The qualities of proposed LEM variants were validated in a series of experiments in static and dynamic enviroment. The results have shown that the metod has better results with smaller group sizes. When compared to the Estimation of Distribution Algorithm, the LEM variants achieve comparable or better values faster. However, the LEM variant which combined the AdaBoost approach with the SVM approach had the best overall performance.
|
|
|
Face-Swap Camera for Android
Škorňok, Petr ; Páldy, Alexander (referee) ; Szentandrási, István (advisor)
The aim of this thesis is to explore existing possibilities of face detection on mobile devices supporting operating system Android and based on these findings create a face swap application with the input from the camera. The overall application is designed with an effort to reach the highest speed possible while processing the images. Implementation is tested from the view of the user and also from the point of program's speed and functionality.
|
|
|
Detection of racist symbols in pictures
Klapal, Matěj ; Říha, Kamil (referee) ; Povoda, Lukáš (advisor)
Goal of this thesis is detector of racist symbols from the picture using functions from the open source library OpenCV. Text also summarizes description of basic processes of image processing via computers. This text contains descriptions of some methods from the library allowing us to train and afterwards detect and localize requested object. This text also compares accuracy of detection using Haar-like features, Local Binary Patterns (LBP) and histogram of oriented gradients. Text also summarizes results of a test of detection for three supported symbols, swastika, signs of SS and triskelion.
|
|
|
Scala Programming Language and Its Use for Data Analysis
Kohout, Tomáš ; Bartík, Vladimír (referee) ; Zendulka, Jaroslav (advisor)
This thesis deals with comparing the Scala programming language with other commonly used languages for data analysis. These languages are evaluated on the basis of the following categories: data manipulation and visualization, machine learning and concurent processing capabilities. The evaluation then shows the strengths and weaknesses of Scala. The strengths will be demonstrated on application for email categorization.
|
|
|
Exploitation of Graphics Processor as Accelerator - OpenCL Technology
Hrubý, Michal ; Jošth, Radovan (referee) ; Zemčík, Pavel (advisor)
This work deals with the OpenCL technology and its use for the task of object detection. The introduction is devoted to description of OpenCL fundamentals, as well as basic theory of object detection. Next chapter of the work is analysis, with design proposal which takes into consideration the possibilities of OpenCL. Further, there's description of implementation of detection application and experimental evaluation of detector's performance. The last chapter summarizes the achieved results.
|
|
|
Face detection in image
Malach, Tobiáš ; Přinosil, Jiří (referee) ; Mézl, Martin (advisor)
This paper presents an overview of face detection methods. Keywords and basic principles of classification of images and it’s parts are explained. Significant part of this paper is occupied with presentation of Viola-Jones detector and it’s implementation in Matlab. Detector Viola-Jones ranks among the most used methods for face detection in practice, which was the reason for detailed analysis and subsequent implementation. Detector is theoretically described, basic steps of algorithm and training algorithm are discussed. Based on theoretical analysis, detector is implemented in Matlab. Properties of implemented detector are objectively evaluated and compared with of two different implementations.
|
|
|
Hardware acceleration of object detection in images
Musil, Petr ; Chalmers, Alan (referee) ; Kadlec, Jiří (referee) ; Zemčík, Pavel (advisor)
V dnešní době je patrný nárůst počtu kamer a dohledových systémů ve veřejném prostoru. Množství informací které tato zařízení produkují je enormní a není v lidských silách je všechny vyhodnotit a interpretovat. Použití výpočetních technologií je nezbytné. Moderní algoritmy počítačového vidění již dosahují skvělých výsledků, jejich širšímu použití v praxi zatím brání nízký výkon zařízení a vysoké požadavky na výpočetní zdroje a energii. Jednou z možností je využití vysokého paraelního výkonu FPGA pro efektivní zpracování těchto algoritmů. Cílem této disertační práce je představit navržené metody optimalizace detektoru objektů v obraze běžících na FPGA. Tyto detektory využívají boostovatelné soft kaskády klasifikátorů spolu s lokálními obrazovými příznaky, které slouží jako slabé klasifikátory. Navržené postupy využívají sekvenční vyhodnocení slabých klasifikátoru. Pro zvýšení výkonu detekce je vyhodnocováno současně více pozic v obraze. Je navržen nový přístup pro detekci objektů různé velikosti nevyžadující externí paměť. Vytvořené detektory byly experimentálně ověřeny na úlohách detekce obličejů a poznávacích značek automobilů. Dosažená výsledky překonávají současný stav poznání, umožňují vytvořit detektory objektů s vyšším detekčním výkonem, lepším poměrem výkonu a spotřebovaných zdrojů FPGA a s lepší přesností detekce.
|
|
|
Face Detection in Video
Hypský, Roman ; Svoboda, Pavel (referee) ; Polok, Lukáš (advisor)
This bachelor's thesis describes detection of faces in video. Methods for human skin detection are described, both using the parametric approach and using the Gaussian distribution. The following is an analysis of face detection methods using the invariant features and using the machine learning methods. A particular emphasis was placed on face detection using the Viola-Jones detector using AdaBoost algorithm. The thesis also describes a practical implementation of the face detector, which was implemented in C++ using OpenCV library. Finally, the described methods are evaluated and the future project development is analyzed.
|
|
|
Pattern Recognition Using AdaBoost
Wrhel, Vladimír ; Hradiš, Michal (referee) ; Herout, Adam (advisor)
This paper deals about AdaBoost algorithm, which is used to create a strong classification function using a number of weak classifiers. We familiarize ourselves with modifications of AdaBoost, namely Real AdaBoost, WaldBoost, FloatBoost and TCAcu. These modifications improve some of the properties of algorithm AdaBoost. We discuss some properties of feature and weak classifiers. We show a class of tasks for which AdaBoost algorithm is applicable. We indicate implementation of the library containing that method and we present some tests performed on the implemented library.
|
|
| |
guest ::
