We invoke an auto-regressive IIR inverse model for convolutive ICA and derive expressions for the likelihood and its gradient. We argue that optimization will give a stable inverse. When there are more sensors than sources the mixing model parameters are estimated in a second step by least squares estimation. We demonstrate the method on synthetic data and finally separate speech and music in a real room recording. © Springer-Verlag 2001.
CITATION STYLE
Dyrholm, M., & Hansen, L. K. (2004). CICAAR: Convolutive ICA with an auto-regressive inverse model. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3195, 594–601. https://doi.org/10.1007/978-3-540-30110-3_76
Mendeley helps you to discover research relevant for your work.