Robust voice activity detection using higher-order statistics in the LPC residual domain.

This paper presents a robust algorithm for voice activity detection (VAD) based on newly established properties of the higher order statistics (HOS) of speech. Analytical expressions for the third and fourth-order cumulants of the LPC residual of short-term speech are derived assuming a sinusoidal m...

Full description

Bibliographic Details
Published in:	IEEE Transactions on speech and audio processing 9, 3 (2001).
Main Author:	Nemer, E.
Format:	Article
Language:	English
Subjects:	Gaussian noise. ITU-T G.729B VAD. LPC prediction error. LPC residual domain. VAD algorithm. Autocorrelation function. Closed-form expressions. Correct classification probability. False classification probability. Flat spectral feature. Fourth-order cumulants. Harmonic content. Higher-order statistics. Kurtosis. Likelihood measure. Low SNR. Noise frames. Periodicity. Phase. Short-term speech. Sinusoidal model. Skewness. Speech detection. Third-order cumulants. Voice activity detection. Voiced speech.