Joint Mixing Vector and Binaural Model Based Stereo Source Separation

Abstract—In this paper the mixing vector (MV) in the statistical mixing model is compared to the binaural cues represented by in-teraural level and phase differences (ILD and IPD). It is shown that the MV distributions are quite distinct while binaural models overlap when the sources are close to each other. On the other hand, the binaural cues are more robust to high reverberation than MV models. According to this complementary behavior we intro-duce a new robust algorithm for stereo speech separation which considers both additive and convolutive noise signals to model the MV and binaural cues in parallel and estimate probabilistic time-frequency masks. The contribution of each cue to the final decision is also adjusted by weighting the log-likelihoods of the cues em-pirically. Furthermore, the permutation problem of the frequency domain blind source separation (BSS) is addressed by initializing the MVs based on binaural cues. Experiments are performed sys-tematically on determined and underdetermined speech mixtures in five rooms with various acoustic properties including anechoic, highly reverberant, and spatially-diffuse noise conditions. The re-sults in terms of signal-to-distortion-ratio (SDR) confirm the ben-efits of integrating the MV and binaural cues, as compared with two state-of-the-art baseline algorithms which only use MV or the binaural cues. Index Terms—Blind source separation, computational auditory scene analysis, reverberation, time-frequency masking. I