X - International Journal of Information Science and Computer Mathematics (Closed Ed TRF)
Volume 1, Issue 2, Pages 97 - 105
(May 2010)
|
|
STUDY ON ALGORITHM OF BLIND SOURCE SEPARATION WITH VARIATIONAL BAYES METHODOLOGY
Dekui Zhao and Lili Gao
|
Abstract:
Aiming at the blind source separation of the time-varying mix matrix and modeling nonstationary signal, based on state-space model and Bayes independent component analysis (ICA), an algorithm with Variational Bayes methodology is present for blind source separation from time-varying mixture. Posterior distribution of Variational Bayes model is studied, and an adaptive correlation computation of model parameters is proposed. The simulation by nonstationary signals shows that the algorithm significantly improves the capability of separating time-varying nonstationary signal, and the waveform of target signal is reconstructed well. |
| Keywords and phrases: blind source separation, Variational Bayes methodology, posterior distribution. |
| Communicated by Kewen Zhao |
| Number of Downloads: 50 | Number of Views: 224 |
|
