Combining pulse-based features for rejecting far-field speech in a HMM-based Voice Activity Detector

Authors:
íscar Varela;Rubén San-Segundo;Luís A. Hernández
Affiliations:
Telephony Platforms Group at Indra S.A., C/Anabel Segura, 7, 28108-Alcobendas, Madrid, Spain;Speech Technology Group at Universidad Politécnica de Madrid, Spain;Applications and Signal Processing Group at Universidad Politécnica de Madrid, Spain
Venue:
Computers and Electrical Engineering
Year:
2011

Citing 6
Cited 2

Induction of fuzzy decision trees

Fuzzy Sets and Systems
Spectrum sensing algorithms for primary detection based on reliability in cognitive radio systems

Computers and Electrical Engineering
Performance evaluation of artificial neural network-based learning schemes for cognitive radio systems

Computers and Electrical Engineering
A PSO-based weighting method for linear combination of neural networks

Computers and Electrical Engineering
A new error concealment method for consecutive frame loss based on CELP speech

Computers and Electrical Engineering
A study on speaker adaptation of the parameters of continuousdensity hidden Markov models

IEEE Transactions on Signal Processing

Voice activity detection algorithm using nonlinear spectral weights, hangover and hangbefore criteria

Computers and Electrical Engineering
A study of voice activity detection techniques for NIST speaker recognition evaluations

Computer Speech and Language

Quantified Score

Hi-index	0.00

Visualization

Abstract

Nowadays, several computational techniques for speech recognition have been proposed. These techniques suppose an important improvement in real time applications where speaker interacts with speech recognition systems. Although researchers proposed many methods, none of them solve the high false alarm problem when far-field speakers interfere in a human-machine conversation. This paper presents a two-class (speech and non-speech classes) decision-tree based approach for combining new speech pulse features in a VAD (Voice Activity Detector) for rejecting far-field speech in speech recognition systems. This Decision Tree is applied over the speech pulses obtained by a baseline VAD composed of a frame feature extractor, a HMM-based (Hidden Markov Model) segmentation module and a pulse detector. The paper also presents a detailed analysis of a great amount of features for discriminating between close and far-field speech. The detection error obtained with the proposed VAD is the lowest compared to other well-known VADs.