Abstract

Besides noise reduction an important objective of binaural speech enhancement algorithms is the preservation of the binaural cues, i.e. the Interaural Level Difference and the Interaural Time Difference of all sound sources. Recently, extensions of the binaural Multi-channel Wiener filter (MWF) have been presented, which aim to preserve the binaural cues of the residual noise component. However, since these algorithms aim to preserve the Interaural Transfer Function (ITF), they are well-suited only for directional noise sources but not for, e.g. spatially isotropic noise, which can not be fully described by the ITF. In this paper, we present an extension of the binaural MWF, aiming to preserve the Interaural Coherence of the residual noise component. Experimental results using spatially isotropic noise show that the proposed algorithm yields a good preservation of the Interaural Coherence without significantly degrading the output SNR compared to the binaural MWF and the binaural MWF with ITF preservation.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call