Blind source separation combining independent component analysis and beamforming
EURASIP Journal on Applied Signal Processing
Speech Enhancement
Hi-index | 0.00 |
In this paper, we propose an effective method for blind speech separation of convolutive mixtures in the frequency domain. The main difficulty in a frequency approach is the permutation problem. In the proposed method, we use two previous approaches to solve permutation ambiguity in the two-channel model: direction of arrival (DOA) estimation approach and assignment problem (AP) approach. We discuss the advantages and disadvantages of the two approaches, and integrate them to use the both advantages. Computer simulation experiments with speech data are presented to illustrate the proposed method.