BLIND SEPARATION OF BINAURAL SOUND MIXTURES USING SIMO-ICA WITH SELF-GENERATOR FOR INITIAL FILTER (TueAmPO2)

Author(s) : Tomoya Takatani (Nara Institute of Science and Technology, Japan) Satoshi Ukai (Nara Institute of Science and Technology, Japan) Tsuyoki Nishikawa (Nara Institute of Science and Technology, Japan) Hiroshi Saruwatari (Nara Institute of Science and Technology, Japan) Kiyohiro Shikano (Nara Institute of Science and Technology, Japan)

Abstract : In this paper, we address the blind separation problem of binaural mixed signals, and propose a novel blind separation method using Single-Input-Multiple-Output-model-based independent component analysis (SIMO-ICA) with a self-generator (SG) for the initial filter. SIMO-ICA which has been proposed by the authors can separate mixed signals, not into monaural source signals but into SIMO-model-based signals from independent sources as they are at the microphones. Although this attractive feature of SIMO-ICA is beneficial to the binaural sound separation, SIMO-ICA has a serious drawback in its high sensitivity to the initial settings of the separation filter. In the proposed method, the SG functions as the preprocessor of SIMO-ICA, and it can provide a valid initial filter for SIMO-ICA. To evaluate its effectiveness, binaural sound separation experiments are carried out under a reverberant condition. The experimental results reveal that the separation performance of the proposed method is superior to those of conventional methods.

Menu