11th Speech in Noise Workshop, 10-11 January 2019, Ghent, BE

Comparison of binaural MVDR-based beamforming algorithms using an external microphone

Nico Gößling, Simon Doclo(a)
Department of Medical Physics and Acoustics and Cluster of Excellence Hearing4All, University of Oldenburg, Germany

(a) Presenting

Besides reducing undesired sound sources, an important objective of a binaural noise reduction algorithm is to preserve the spatial impression of the acoustic scene for the hearing aid user, such that no mismatch between acoustic and visual information occurs and the binaural hearing advantage can be exploited. Although the binaural minimum variance distortionless response (MVDR) beamformer is able to preserve the binaural cues of the target speaker, it distorts the binaural cues of the background noise. Hence, several extensions have been proposed, aiming at preserving the binaural cues of the background noise [1]. Because the performance of noise reduction algorithms is partly limited by the physical design, requiring the microphones to be integrated into the hearing devices, the usage of one or more external microphones that are spatially separated from the head-mounted microphones has been recently explored [2-3]. It has been shown that an external microphone enables to improve both the noise reduction performance as well as the binaural cue preservation performance.

In this contribution, we present a comparison of several binaural MVDR-based beamforming approaches that make use of an external microphone. First, the external microphone is merely used to provide an estimate of acoustic variables (e.g., relative transfer functions) that are required for the binaural beamforming algorithms. Second, the external microphone is used in conjunction with the head-mounted microphones as an additional input signal. Using realistic recordings of a moving target speaker in a reverberant room the performance of the considered binaural beamforming approaches is compared in terms of noise reduction performance (using speech-intelligibility-weighted SNR) and binaural cue preservation (using reliable ITD and ILD cues from an auditory model).

[1] D. Marquardt, S. Doclo (2018). “Interaural Coherence Preservation in Binaural Hearing Aids using Partial Noise Estimation and Spectral Postfiltering,” IEEE/ACM Trans. Audio, Speech and Language Processing, 26(7):1257-1270.
[2] J. Szurley, A. Bertrand, B. Van Dijk, M. Moonen (2016). “Binaural noise cue preservation in a binaural noise reduction system with a remote microphone signal,” IEEE/ACM Trans. Audio, Speech and Language Processing, 24(5):952–966.
[3] N. Gößling, S. Doclo (2018). “RTF-Based Binaural MVDR Beamformer Exploiting an External Microphone in a Diffuse Noise Field,” in Proc. ITG Conference on Speech Communication, Oldenburg, Germany, Oct. 2018, pp. 106-110.

Last modified 2018-12-08 00:23:30