CICAAR: Convolutive ICA with an Auto-Regressive Inverse Model |
Mads Dyrholm, Lars Kai Hansen
|
| Abstract | 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. |
| Keywords | Convolutive ICA AR speech deconvolution maximum likelihood |
| Type | Conference paper [With referee] |
| Conference | Independent Component Analysis and Blind Signal Separation |
| Editors | Carlos G. Puntonet and Alberto Prieto |
| Year | 2004 Month September Vol. 3195 pp. 594-601 |
| Publisher | Springer |
| Series | Lecture Notes in Computer Science |
| ISBN / ISSN | 3-540-23056-4 |
| Electronic version(s) | [pdf] |
| BibTeX data | [bibtex] |
| IMM Group(s) | Intelligent Signal Processing |