Efficient blind estimation of subband reverberation time from speech in non-diffuse environments
Salomon Diether, Lukas Bruderer, Andreas Streich, Hans-Andrea Loeliger
Abstract
To respond to reverberation effectively, modern speech processing tasks like signal enhancement in digital hearing aids or distant-talking speech recognition often require precise knowledge of the acoustic situation. A common measure is the (frequency-dependent) reverberation time T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> . Explicit measurement using sound excitations is not possible for most applications, therefore blind estimation of subband T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> from speech signals is employed. Previous approaches are either limited to T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> between 0...1.5 s, or require long speech signals and costly computation. We propose an efficient algorithm that estimates T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> up to 6 s from short speech signals of no more than 20 s. In experiments with real room impulse responses (RIR) the algorithm exhibits state-of-the-art performance for T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> up to 1.5 s and superior performance for longer T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> .
BibTeX
@inproceedings{icassp2015_efficientblindes,
title = {Efficient blind estimation of subband reverberation time from speech in non-diffuse environments},
author = {Salomon Diether and Lukas Bruderer and Andreas Streich and Hans-Andrea Loeliger},
booktitle = {ICASSP 2015},
year = {2015}
}