Title :
Robust voice activity detection algorithm using spectrum estimation and dynamic thresholding
Author :
Zenteno, Efrain ; Sotomayor, Manuel
Author_Institution :
Telecommun. Eng., Univ. Catolica San Pablo, Arequipa, Peru
Abstract :
An algorithm for voice activity detection (VAD) is presented in this paper, it uses the classical short time Fourier transform (STFT) followed by dynamic thresholding over the noise floor which allows accurate determination for voiced/unvoiced frame. This algorithm has also been implemented in a DSP plataform to evaluate its performance in real time, where results have shown that such method presents accurate voiced/unvoiced determination under unknown signal to noise ratio (SNR) conditions.
Keywords :
Fourier transforms; signal detection; speech processing; DSP plataform; digital speech processing; dynamic thresholding; short time Fourier transform; spectrum estimation; voice activity detection algorithm; Detection algorithms; Fourier transforms; Frequency; Robustness; Signal to noise ratio; Spectral analysis; Spectrogram; Speech analysis; Speech enhancement; Speech processing;
Conference_Titel :
Communications, 2009. LATINCOM '09. IEEE Latin-American Conference on
Conference_Location :
Medellin
Print_ISBN :
978-1-4244-4387-1
Electronic_ISBN :
978-1-4244-4388-8
DOI :
10.1109/LATINCOM.2009.5304908
