Title :
Noise robust voice activity detection using normal probability testing and time-domain histogram analysis
Author :
Ghaemmaghami, Houman ; Dean, David ; Sridharan, Sridha ; McCowan, Iain
Author_Institution :
Speech & Audio Res. Labs., Queensland Univ. of Technol., Brisbane, QLD, Australia
Abstract :
This paper presents a method of voice activity detection (VAD) suitable for high noise scenarios, based on the fusion of two complementary systems. The first system uses a proposed non-Gaussianity score (NGS) feature based on normal probability testing. The second system employs a histogram distance score (HDS) feature that detects changes in the signal through conducting a template-based similarity measure between adjacent frames. The decision outputs by the two systems are then merged using an open-by-reconstruction fusion stage. Accuracy of the proposed method was compared to several baseline VAD methods on a database created using real recordings of a variety of high-noise environments.
Keywords :
speech processing; time-domain analysis; histogram distance score feature; noise robust voice activity detection; nonGaussianity score feature; normal probability testing; open-by-reconstruction fusion stage; speech processing; template-based similarity measure; time-domain histogram analysis; Databases; Feature extraction; Histograms; Noise robustness; Signal to noise ratio; Speech analysis; Speech enhancement; Speech processing; System testing; Time domain analysis; decision fusion; histogram analysis; normal probability; voice activity detection;
Conference_Titel :
Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference on
Conference_Location :
Dallas, TX
Print_ISBN :
978-1-4244-4295-9
Electronic_ISBN :
1520-6149
DOI :
10.1109/ICASSP.2010.5495612