Frequency-Domain Blind Source Separation (BSS) with Parametric Pearson Distribution Modeling

In frequency-domain blind source separation (BSS) for speech signals using independent component analysis (ICA), a practical parametric Pearson distribution system is employed to model the statistical distribution of frequency-domain source components, improving separation accuracy through optimized probability density function matching.