An EM Algorithm for Independent Component Analysis Using an AR-GGD Source Model

A maximum likelihood blind source separation algorithm is developed. The temporal dependencies are explained by assuming that each source is an autoregressive (AR) process and the distribution of the associated i.i.d. innovations process is described by generalized Gaussian distribution (GGD), which can fit a broader range of statistical distributions by varying the value of the steepness parameter α. Unlike most maximum likelihood methods the proposed algorithm takes into account both spatial and temporal information. Optimization is performed using the Expectation-Maximization method, and the source model is learned alone with the demixing parameters.