Beta 1


Title Probabilistic Speech Detection
Author Jacobsen, Daniel J.
Supervisor Larsen, Jan (Department of Informatics and Mathematical Modeling, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Institution Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
Thesis level Master's thesis
Year 2003
Abstract This thesis deals with the detection of speech in signals that may contain very different noise types, referred to as the 'Voice Activity Detection' (VAD) problem. The signals consist of sections of noise only and sections of speech and noise in an additive mixture; convolutive mixtures are not addressed. Two different probabilistic methods are developed to solve the VAD problem. One is a discriminant-function based method in which a linear network with a single logistic output is trained to output the probability of speech presence from a given sound signal. The other is based on modelling of class-conditional probability densities, using Independent Component Analysis (ICA) methods. The algorithms are tested extensively and comparisons are made between them. They are also compared to an industry standard VAD algorithm, namely that of the the ITU-T G.729B recommendation and one other VAD. The results show the crucial importance of considering the type of noise present with the speech for obtaining robust speech detection and that for certain noise types, performance can be bettered with the developed VAD algorithms.
Imprint Department of Informatics and Mathematical Modeling, Technical University of Denmark, DTU : DK-2800 Kgs. Lyngby, Denmark
Keywords machine learning; classification; voice activity detection; linear networks; independent component analysis; receiver operating characteristics
Fulltext
Original PDF imm2551.pdf (1.93 MB)
Admin Creation date: 2006-06-22    Update date: 2012-12-20    Source: dtu    ID: 58623    Original MXD