Akustična podobnost besed pri avtomatskem razpoznavanju govora

diplomsko delo univerzitetnega študijskega programa

Nina Vodopivc (Author), Andrej Žgank (Mentor), Mirjam Sepesy Maučec (Co-mentor)

Abstract

V diplomskem delu smo izdelali program, ki na izhodu poda ocene akustične podobnosti besed na podlagi Levenshteinove razdalje. Za pisanje programske kode smo uporabili prosto dostopno programsko okolje NetBeans IDE 6.9.1. Uporabili smo tudi prosto dostopno orodje HTK, kjer smo na realnem modelu razpoznavalnika govora s pripravo dveh testnih scenarijev izračunali rezultate razpoznavanja govora za dva različna nabora besed. Za vsak nabor smo izvedli več eksperimentov. Rezultate obeh scenarijev smo primerjali in ocenili vpliv akustične podobnosti besed na uspešnost razpoznavanja govora.

Keywords

komunikacija človek stroj;avtomatski razpoznavalniki govora;akustično modeliranje;

Data

Language:	Slovenian
Year of publishing:	2011
Source:	Maribor
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[N. Vodopivc]
UDC:	621.391:004.93(043.2)
COBISS:	15758614
Views:	1468
Downloads:	89
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Automatic speech recognition and acoustic confusability of words
Secondary abstract:	In diploma work we have created a program, which calculates an estimation of acoustic confusability of words based on Levenshtein distance. An open source software environment NetBeans IDE 6.9.1 was used for developing a program code. We also used a freeware HTK toolkit, which provided us results of recognition for two different sets of words on realistic model of automatic speech recogniser. For each set of words different experiments were run. We compared results from both scenarios, described them and estimated influence of acoustic confusability of words on accuracy of speech recognition.
Secondary keywords:	human-computer interaction;automatic speech recognition;acoustic modeling;
URN:	URN:SI:UM:
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Telekomunikacije
Pages:	VII, 42 f.
Keywords (UDC):	applied sciences;medicine;technology;uporabne znanosti;medicina;tehnika;engineering;technology in general;inženirstvo;tehnologija na splošno;mechanical engineering in general;nuclear technology;electrical engineering;machinery;strojništvo;electrical engineering;elektrotehnika;science and knowledge;organization;computer science;information;documentation;librarianship;institutions;publications;znanost in znanje;organizacije;informacije;dokumentacija;bibliotekarstvo;institucije;publikacije;prolegomena;fundamentals of knowledge and culture;propaedeutics;prolegomena;splošne osnove znanosti in kulture;computer science and technology;computing;data processing;računalniška znanost in tehnologija;računalništvo;obdelava podatkov;application-oriented computer-based techniques;računalniške tehnike za namensko rabo;aplikativno usmerjene računalniško podprte tehnike;pattern information processing;obdelava informacij v vzorcih;
ID:	1015646