A Corpus-Based Evaluation of Beamforming Techniques and Phase-Based Frequency Masking

Abstract

Beamforming is a type of audio array processing techniques used for interference reduction, sound source localization, and as pre-processing stage for audio event classification and speaker identification. The auditory scene analysis community can benefit from a systemic evaluation and comparison between different beamforming techniques. In this paper, five popular beamforming techniques are evaluated in two different acoustic environments, while varying the number of microphones, the number of interferences, and the direction-of-arrival error, by using the Acoustic Interactions for Robot Audition (AIRA) corpus and a common software framework. Additionally, a highly efficient phase-based frequency masking beamformer is also evaluated, which is shown to outperform all five techniques. Both the evaluation corpus and the beamforming implementations are freely available and provided for experiment repeatability and transparency. Raw results are also provided as a complement to this work to the reader, to facilitate an informed decision of which technique to use. Finally, the insights and tendencies observed from the evaluation results are presented.

Postfilter for Dual Channel Speech Enhancement Using Coherence and Statistical Model-Based Noise Estimation

“…where γ = Φ x Φ n is the SNR at the beamformer output, which is estimated using the statistical model-based noise PSD estimate in (8)…”

“…Over the past decades, there has been a growing demand for speech enhancement using microphone arrays in speech processing applications such as automatic speech recognition, mobile communications, and hearing aids [ 1 , 2 , 3 , 4 ]. Multichannel speech enhancement aims to reduce the additive noise and improve the quality of the speech signals obtained by multiple microphones placed in a variety of acoustic environments [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 ]. In many multichannel speech enhancement systems, beamforming algorithms, such as the minimum-variance distortionless-response (MVDR) beamformer [ 11 ] and the general transfer function generalized sidelobe canceler (TF-GSC) [ 12 , 13 ], have been employed to extract a desired signal, exploiting spatial information on the location of the sound sources.…”

On the Challenges of Acoustic Energy Mapping Using a WASN: Synchronization and Audio Capture

“…Phase-based Binary Masking (PBM) [ 24 , 25 ] is not an actual beamformer, but it has a similar application for the acoustic mapping. This technique employs the first stage of the DAS beamformer, such that the information arriving from a given steered position is aligned in all nodes.…”

Target Selection Strategies for Demucs-Based Speech Enhancement

“…There are several beamforming techniques that can be used for this purpose, such as minimum variance distortionless response (MVDR) [33,34] or generalized sidelobe canceller (GSC) [35,36]. However, it has been shown [37] that a simple phase-based frequency masking (PFM) [38] is able to obtain a good amplification of the speech source at a given location while still being able to run in an online manner.…”

“…Small values of σ may provide a better isolation of the target source, but this occurs at the expense of increased sensitivity to variation in the target location. Values between 10 • and 30 • have been shown to provide a good balance between these two issues [37].…”