Thesis Details
Agreements and Disagreements between Automatic and Human Speaker Recognition
This thesis deals with the problem of speaker recognition. The term is defined and individual methods related to it are described. The aim of the work is to point out the agreements and disagreements between human and automatic speaker recognition. At the beginning of the thesis, theoretical knowledge from both mentioned areas are described, i.e., what aspects of human speech do people, or automatic system respectively, focus on. Then, several experiments are performed to compare these two methods. These experiments are evaluated in such a way, that it is possible to observe which tasks can be better solved by people, in order to use the knowledge to improve the function of the automated system. At the end of the thesis, such an attempt to improve the automatic system is demonstrated and tested. Testing was successful and higher evaluation accuracy could be observed. Thus, such a result can be used in other research to allow further development in the field of automatic speaker recognition.
recognition, speaker, human, automatic, voice, sound, fonetics, fonology, forensics, system, model, identification, verification, detection, idiosyncratic, analysis
Hliněná Dana, doc. RNDr., Ph.D. (DMAT FEEC BUT), člen
Jaroš Jiří, doc. Ing., Ph.D. (DCSY FIT BUT), člen
Orság Filip, Ing., Ph.D. (DITS FIT BUT), člen
Rychlý Marek, RNDr., Ph.D. (DIFS FIT BUT), člen
@bachelorsthesis{FITBT21943,
author = "Jakub Valenta",
type = "Bachelor's thesis",
title = "Agreements and Disagreements between Automatic and Human Speaker Recognition",
school = "Brno University of Technology, Faculty of Information Technology",
year = 2019,
location = "Brno, CZ",
language = "english",
url = "https://www.fit.vut.cz/study/thesis/21943/"
}