|
|
Reconstruction of Damaged Parts of Fingerprint Image Using Neural Nets
Bobocký, Boris ; Dyk, Tomáš (referee) ; Kanich, Ondřej (advisor)
In this paper, I propose a method for reconstructing damaged fingerprints using generative adversarial networks (GANs), implemented with Python and the PyThorch library. I have trained a specific GAN model on a dataset of approximately twenty thousand prints, created with Anguli and other damage simulation tools. This approach produced excellent results and could have wide application in biometric systems. This work highlights the potential of deep learning in the fields of image reconstruction and biometrics.
|
|
|
Preparation and characterization of nanostructured III-V semiconductor materials
Maniš, Jaroslav ; Kostelník,, Petr (referee) ; Hospodková,, Alice (referee) ; Šikola, Tomáš (advisor)
Předkládaná dizertační práce se zabývá výrobou a analýzou gallium nitridových (GaN) nanostruktur ve třech odlišných formách. V prvním případě byl zkoumám trojdimenzionální GaN ve formě nanokrystalů rostených na grafenu. Nanokrystaly byly připraveny s využitím techniky droplet epitaxy, která mimo jiné umožňuje růst nanostruktur za nízké teploty substrátu (T = 200°C). Studium se zaměřovalo jak na charakterizaci kvality připravených nanokrystalů, tak na statistický popis růstu. V dalším kroku byly připravené struktury využity pro výrobu fotodektoru citlivého na ultrafialové světlo. Výroba fotodektoru a jeho úspěšné použití slouží jako základ pro navazující výzkum. Ve druhém případě byly studovány dvoudimenzionální GaN nanostruktury, které byly rovněž připraveny za nízké teploty křemíkového substrátu. Následná analýza se soustředila na popis krystalové struktury a prvkovou analýzu, neboť byly takovéto struktury pozorovány vůbec poprvé. Další rozvoj možností přípravy těchto nanostruktur je předmětem navazujícího výzkumu. Ve třetím případě byly zkoumány jednodimenzionální GaN nanodráty připravené na safírovém substrátu. Účelem tohoto projektu bylo získání datasetu pro ověření teoretického modelu, který popisuje růst horizontálních nanodrátů. Na základě sběru a analýzy dat se podařilo modelovat růstovou dynamiku GaN nanodrátů, která byly v souladu s teoretickým modelem.
|
|
|
The photoluminescence properties measurement of ultrathin films
Metelka, Ondřej ; Mach, Jindřich (referee) ; Šamořil, Tomáš (advisor)
The thesis briefly describes the principles and types of luminescence. In the first following research of study is also discussed the equipment which is applicable to photoluminescence experiments, including the arrangement. The second research focuses on the influence of the properties of gallium nitride (GaN) (ultra) thin films and other structures prepared by various ways on shape of photoluminescence spectra. The paperwork also describes the further optimization of photoluminescent apparatus used for the measurement of photoluminescence spectrum in the UV light radiation which is located at the Institute of Physical Engineering at the Technical University. The extension of measurements at low temperatures (design and construction of its own cryostat) is added. The conclusion concernes the test measurements to determine the effect of various settings of the apparatus on the resulting measured photoluminescence spectrum.
|
|
|
Development and Application of an UHV Equipment for Deposition of Thin Films (Atomic and Ion Systems)
Mach, Jindřich ; Čech, Vladimír (referee) ; Lencová, Bohumila (referee) ; Šikola, Tomáš (advisor)
In the thesis the development of two equipment for preparation of ultrathin films under ultrahign vacuum conditions (UHV) is discussed. Here, additionally to a brief description of theoretical principles, more details on the design of these units are given. In the first part the design of a thermal source of oxygen or hydrogen atomic beams is discussed. Further, a design and construction of an ion–atomic beam source for ion-beam assisted deposition of thin films is detailed. The source combines the principles of an efusion cell and electron-impact ion beam source generating ions of (30 – 100) eV energy. The source has been successfully applied for the growth of GaN on the Si(111) 7x7 substrate under room temperature.
|
|
|
Robust Audio Dereverberation and Denoising
Košina, Simon ; Skácel, Miroslav (referee) ; Szőke, Igor (advisor)
The goal of this thesis was to create a speech enhancement and dereverberation model for audio recordings coming from aircraft VHF communication. First, the thesis covers some theoretical grounds of machine learning and types of neural networks commonly used in such scenarios. Following is a description of the used framework, datasets and the implementation itself. Last chapters are focused on the performed experiments and their evaluation. At the end we talk about the future work that can be done in order to further improve the achieved results.
|
|
|
Generating training data with neural networks
Ševčík, Pavel ; Kolář, Martin (referee) ; Hradiš, Michal (advisor)
The aim of this thesis was to prepare a training data set for traffic sign detection using generative neural networks. The solution uses a modified U-Net architecture and several experiments with the application of styles using AdaIN layers as in the StyleGAN model have been conducted. By extending the real GTSDB data set with the generated images, mean average precision of 80.36 % has been achieved, which yields an improvement of 19.27 % compared to the mean average precision of the detection model trained on real data only.
|
|
|
Image Super-Resolution Using Deep Learning
Mojžiš, Tomáš ; Beran, Vítězslav (referee) ; Španěl, Michal (advisor)
The aim of this thesis is to create a deep neural net capable of super-resolution on images acquired by electron microscopes. The thesis consists of two parts - finding appropriate data and creating a dataset for the super-resolution task and designing a neural net architecture capable of solving the super-resolution task. Within the thesis, two datasets comprised of images acquired by electron microscopes were created. The datasets differ in the approach to data augmentation. They allow to train a neural network which fulfills the super-resolution task. To solve this task, two U-Net based and one GAN based architecture were trained. The resolution of images was upscaled by a factor of two and four. The best artificially upscaled images were created by neural network Real-ESRGAN. The values of metrics were not higher than the tested interpolation method, but the images seem more visually pleasing especially when they were upscaled four times. Thanks to this thesis, two datasets were created allowing to train other possible neural network architectures to improve the quality of the artificially upscaled images. The neural networks trained in this thesis can be utilized in the process of acquiring higher quality data from low resolution electron microscope images.
|
|
|
UV detection by Grafen/GaN structures
Kostka, Marek ; Piastek, Jakub (referee) ; Mach, Jindřich (advisor)
This bachelor thesis covers fabrication as well as the study of behavior of UV sensor based on galium nitride structures on graphene substrate. The sensor takes advantage of high mobility of charge carriers in graphene and UV selective sensitivity of GaN nanoctystals. This combination exhibits a potential for being a UV sensor because of the unique properties brought by both materials. The first part of this work describes graphene, galium nitride and graphene/GaN heterostructure with their properties and fabrication processes. The second part includes full description of already used fabrication process and characterization of the optical and electrical behavior using photoluminiscence and transport properties measurements.
|
|
|
Development of Atomic- and Ion Beam Sources
Šamořil, Tomáš ; Lencová, Bohumila (referee) ; Mach, Jindřich (advisor)
The objective of this master thesis was to provide the optimization of an ion-atom beam source for the improvement of its properties. The improvement of the parameters increases the efficiency of the source during the deposition of gallium nitride ultrathin films (GaN) being important in microeletronics and optoelectronics. After optimization, the depositions of GaN ultrathin films on Si(111) 7x7 at lower temperatures (
|
|
|
Learning the Face Behind a Voice
Kyjonka, Mojmír ; Matějka, Pavel (referee) ; Plchot, Oldřich (advisor)
This thesis deals with face reconstruction based on voice. The state of the art of this problem is investigated and model for such problem is trained. Model used in this thesis is based on the work "Reconstructing faces from voices" which architecture is based on Generative Adversarial Network (GAN). In this work, we used VGGFace and VoxCeleb datasets, and additionally, we created a small audiovisual dataset of Czech speakers. This work was implemented using the Python scripting language and PyTorch library.
|
guest ::
