Comparative Study of Single-Channel Algorithms for Blind Reverberation Time Estimation

Löllmann H, Brendel A, Kellermann W (2019)

Publication Language: English

Publication Type: Conference contribution, Conference Contribution

Publication year: 2019

Event location: Aachen DE

Abstract

Knowledge about the reverberation time (T60) is exploited by numerous algorithms, e.g., for automatic speech recognition, speech dereverberation, or as a feature for acoustic scene classification. This contribution provides a comprehensive survey about various state-of-the-art methods to estimate the reverberation time blindly from a noisy and reverberant speech signal and compares their estimation performance for different acoustic scenarios. The evaluation considers different acoustic conditions regarding the signal-to-noise ratio (SNR), direct-to-reverberation ratio (DRR) and T60, and uses databases with room impulse responses measured in different acoustic enclosures. The findings of the Acoustic Characterisation of Environments (ACE) Challenge are extended by taking into account algorithms and acoustic scenarios which are not considered in this benchmarking campaign, and by using a much larger set of reverberant speech signals.

Authors with CRIS profile

Heinrich Löllmann Professur für Signalverarbeitung Andreas Brendel Lehrstuhl für Multimediakommunikation und Signalverarbeitung Walter Kellermann Professur für Signalverarbeitung

How to cite

APA:

Löllmann, H., Brendel, A., & Kellermann, W. (2019). Comparative Study of Single-Channel Algorithms for Blind Reverberation Time Estimation. In Proceedings of the International Congress on Acoustics (ICA). Aachen, DE.

MLA:

Löllmann, Heinrich, Andreas Brendel, and Walter Kellermann. "Comparative Study of Single-Channel Algorithms for Blind Reverberation Time Estimation." Proceedings of the International Congress on Acoustics (ICA), Aachen 2019.

BibTeX: Download