Parametric-Pearson-based independent component analysis for frequency-domain blind speech separation

Separation performance is improved in frequency-domain blind source separation (BSS) of speech with independent component analysis (ICA) by applying a parametric Pear-son distribution system. ICA adaptation rules include a score function determined by approximated source distribution, and better approximation improves separation per-formance. Previously, conventional hyperbolic tangent (tanh) or generalized Gaussian distribution (GGD) was uniformly applied to the score function for all frequency bins, despite the fact that a wideband speech signal has different distributions at different frequencies. To obtain better score functions, we propose the integration of a parametric Pear-son distribution system with ICA learning rules. The score function is estimated by using appropriate Pearson distribu-tion parameters for each frequency bin. We consider three estimation methods with Pearson distribution parameters and conduct separation experiments with real speech sig-nals convolved with actual room impulse responses. Conse-quently, the signal-to-interference ratio (SIR) of the pro-posed methods significantly improve over 3 dB compared to conventional methods.