Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Blind single-channel dereverberation using a recursive maximum-sparseness-power-prediction-model

: Herzog, A.; Habets, E.A.P.


Saruwatari, H. ; Institute of Electrical and Electronics Engineers -IEEE-:
16th International Workshop on Acoustic Signal Enhancement, IWAENC 2018. Proceedings : 17th-20th September 2018, Tokyo, Japan
Piscataway, NJ: IEEE, 2018
ISBN: 978-1-5386-8151-0
ISBN: 978-1-5386-8150-3
ISBN: 978-1-5386-8152-7
International Workshop on Acoustic Signal Enhancement (IWAENC) <16, 2018, Tokyo>
Fraunhofer IIS ()

Several single-channel speech dereverberation techniques rely on a sufficiently precise estimate of the reverberation time T 60 . Blindly estimating the frequency dependent T 60 remains a challenging task, especially if only a short time section is available for the estimation. In this work, we first review an offline method which estimates both T 60 per frequency as well as the power spectral density (PSD) of the early reverberant speech and then derive a recursive and adaptive version thereof. These estimates can be used to suppress the late reverberation of reverberant speech signals. It is shown that the proposed algorithm can estimate the T 60 with a deviation of approximately ±20% w.r.t. the ground truth value. Moreover, objective quality measures show that the proposed dereverberation algorithm performs slightly better than two other existing methods.