|
|
Implementation of the infrastructure for microphone audio streaming into the cloud
Dvořák, Petr ; Smékal, Zdeněk (referee) ; Zvončák, Vojtěch (advisor)
The task of this bachelor's thesis is creation of simple and functional infrastructure designated for recording and transmission of an audio recording captured by microphone and processed by Raspberry Pi. Final output file is then transfered to cloud storage. The goal of the thesis is making a prototype capable of processing a sound recording into required format and with suitable quality. The work should consider possibility of charging the system with battery, type of used network technology and setup of the cloud storage. Prototype should be simple and functional. Importance is comparing the infrastructure with similar devices on the market.
|
|
|
Identification of specified segments in the audio signal using machine learning
Pařízek, Radim ; Galáž, Zoltán (referee) ; Zvončák, Vojtěch (advisor)
The bachelor thesis deals with the design of a system for the identification of natural environmental sounds in audio recordings. The datasets and models used for this type of tasks are surveyed and their structure is described. A system for the identification of sounds in one layer and in two layers has been proposed for seven selected labels. The classifier used for this system was created by fine-tuning a transformer model from the Hugging Face platform. The results of two training approaches and one identification system were evaluated.
|
|
|
Research of the new augmentation methods for online handwriting
Sigmund, Jan ; Burget, Radim (referee) ; Zvončák, Vojtěch (advisor)
Graphomotor difficulties of school-aged children are characterised by problems in handwriting and drawing and can lead to developmental dysgraphia. Timely clinical diagnosis is critical to provide preventive care. In practice however, it is not feasible on day-to-day basis due to the need for expert staff and the prevalence of difficulties up to 30\%. Machine learning models can serve as an accessible objective tool for evaluating graphomotor functioning. In most cases there is not enough data collected, which results in poor classification performance. Therefore, this thesis focuses on data augmentation of online handwriting. Generating artificial samples is based on recombination of intrinsic mode functions, obtained by empirical mode decomposition. IMFs of health controls, numbering 72, and with graphomotor difficulties, 94 children in total, are calculated. The decomposition is performed specifically on X and Y coordinate time series. IMFs of the same indices of different subjects are randomly interchanged, thus producing a new signal. Then, the graphomotor features of the original and artificial time series are extracted. Only the spatial ones related to the coordinates are selected. Finally, the correlations of the features of the two databases will be analyzed and compared.
|
|
| |
|
|
Comparison of specialized actigraphs with wearable devices in quantitative sleep analysis
Čech, Vladimír ; Zvončák, Vojtěch (referee) ; Mikulec, Marek (advisor)
Actigraphy and smart wearable devices provide similar functionality in quantitative sleep analysis. This work aims to verify how common wearable devices stand in comparison with a specialized actigraph used in clinical practice. The actigraph used in this work was Geneactiv Original and it was used to compare the sleep analysis results of eleven commercially available devices. The measurements for nine of the measured devices lasted for one week. Two devices were measured for six days. During sleep, the device was worn on the same hand as the actigraph, and at the same time a sleep diary was kept, in which data on the time of lying down, the time of awakening and the number of awakenings per night were recorded. After a week of measurement, the data obtained from the actigraph were evaluated by software and then compared with the data provided by the second device. In this work, data of total sleep time, sleep onset latency, number of awakenings during the night and sleep efficiency are compared. Ten of the eleven devices measured higher total sleep time and higher sleep efficiency than the actigraph. These devices measured higher total sleep time by 1.21 % – 12.06 % and measured higher sleep efficiency by 2.86 % - 13.86 %. One device, namely Fossil Sport, measured lower total sleep time by 9.02 % and lower sleep efficiency by 9.13 %. Sleep onset latency was the most distinct parameter. Wearable devices measured higher sleep time by 86.72 % – 1225.95 %. Neither device nor actigraph could reliably determine the number of awakenings during the night unless it was a significant physical activity during the night. From the results of the work, it is not possible to say in general that all commercially available devices would be a reliable substitute for actigraphy.
|
|
|
Prototype system for centralized control of music effects using Android OS
Slezák, Viktor ; Ilgner, Petr (referee) ; Zvončák, Vojtěch (advisor)
This bachelor thesis had been encouraged by the thought of easily operable sound effects from only one place. Guitar pedalboard can be taken in a consideration as an example, where multiple sound effects are operated by the means of small boxes with several turning potentiometers. When a guitarist wants to change the set up of those sound effects, they must manually reconfigure them separately. The aim of this thesis is therefore development of an application and actuator which enable to operate sound effects from one place by a one-click. The goal is to simplify and centralize the control of sound effects, such are guitar effects in pedalboard, into one application. A single-board computer Arduino and additional Bluetooth module HC-05 are being used for the development. Control systems are run on the Android platform. The text itself deals with the realization of the application, which can remotely control actuator for the sound effects regulation, with the programming problematics, communication via Bluetooth platform and the construction of the solution of actuator with regards to printable method FDM.
|
|
|
Automatic speech recordings segmentation tool
Santa, Roman ; Zvončák, Vojtěch (referee) ; Kováč, Daniel (advisor)
Nástroj pre automatickú segmentáciu spracováva nahrávky reči a extrahuje hovorené slovo z nahrávok. Je dôležité, aby pokročilá analýza pracovala iba s rečovými časťami z nahrávky. Nástroj na segmentáciu má ulahčiť spracovanie nahrávok pre analýzu rozdielov medzi hláskami pacientov s parkinsonovou chorobou a tými zdravými. Cieľ tejto práce je navrhnúť a otestovať detektory reči s Google WebRTC detektorom a vybrať ten najvhodnejší detektor reči s minimálnym počtom chýb. Ďalej, vytvoriť nástroj na segmentáciu nahrávok a otestovať rozpoznávanie reči pomocou dynamic time warping. Bola použitá databáza poskytnutá laboratóriom pre analýzu mozgových ochorení. Obsahuje české a maďarské nahrávky s rovnakým počtom mužských a ženských pacientov a aj rovnakým počtom zdravých pacientov a pacientov s parkinsonovou chorobou. Najlepšie výsledky v testoch dosiahol detektor na základe energie reči. Nebol zistený žiaden rozdiel v presnosti detektoru pri spracovaní mužských a ženských nahrávok alebo nahrávok zdravých či chorých pacientov. Nahrávky s nízkym odstupom signálu od šumu boli náročnejšie na spracovanie s frekvenciou chýb od 12%. Na základe výsledkov, bol navrhnutý nový detektor pre spracovanie úplnej nahrávky. Na záver bol testovaný algoritmus pre rozpoznávanie podobnosti reči na základe melovských kepstrálnych koeficientov.
|
|
|
Recognition of music cover versions using Music Information Retrieval techniques
Martinek, Václav ; Zvončák, Vojtěch (referee) ; Kiska, Tomáš (advisor)
This master’s thesis deals with designs and implementation of systems for music cover recognition. The introduction part is devoted to the calculation parameters from audio signal using Music Information Retrieval techniques. Subsequently, various forms of cover versions and musical aspects that cover versions share are defined. The thesis also deals in detail with the creation and distribution of a database of cover versions. Furthermore, the work presents methods and techniques for comparing and processing the calculated parameters. Attention is then paid to the OTI method, CSM calculation and methods dealing with parameter selection. The next part of the thesis is devoted to the design of systems for recognizing cover versions. Then there are compared systems already designed for recognizing cover versions. Furthermore, the thesis describes machine learning techniques and evaluation methods for evaluating the classification with a special emphasis on artificial neural networks. The last part of the thesis deals with the implementation of two systems in MATLAB and Python. These systems are then tested on the created database of cover versions.
|
|
|
Research of new online handwriting features in children with graphomotor difficulties
Gavenčiak, Michal ; Mekyska, Jiří (referee) ; Zvončák, Vojtěch (advisor)
In the Czech Republic, there is currently no objective method to diagnose graphomotor difficulties in children. Ongoing research uses modern digitizers to capture the hand-writing process and quantify its parameters. The first goal of this thesis is to develop software tools to faciliate work with the collected data, such as database validation and writing exercise rating, done by specialists. Another goal of this thesis is to design new on-line handwriting parameters which are then to be analysed on a cohort of school children from 2nd to 4th class of primary school (n=239). The implementation of two desktop programs on the .NET platform is described, among three new quantifying parameters based on the principles of isochrony, two-dimensional cross-correlation, and geometrical centroid. All three parameters show significant correlation (r = [0,2; 0,3])with the HPSQ-C rating in 2nd- and 4th-graders and correlation (𝜌= [0,2; 0,5]) with specialist’s subjective scores in all children from the cohort. The analysis suggests children with graphomotor difficulties struggle with regulating handwriting speed and working memory.
|
|
|
Cancelling noise of magnetic resonance in recordings of speech
Vrba, Filip ; Galáž, Zoltán (referee) ; Zvončák, Vojtěch (advisor)
This thesis deals with the removal of noise in speech recordings that have been recorded in an MRI environment. For this purpose, the Nvidia RTX Voice technology, the VST plug-in module Noisereduce and a self-designed method of subtractive de-noising of recordings are used. A program with a simple graphical interface in Python is implemented within the work to retrieve the recordings and then de-noise them using the proposed methods. The work includes measurements in a magnetic resonance environment with two microphones. The quality of the processed recordings is tested within the program using the STOI (Short-Time Objective Intelligibility Measure) method as well as the subjective analysis method within listening tests.
|
guest ::
RSS feed.