Title :
A New Diagonal Hessian Algorithm for Blind Signal Separation
Author :
ElSabrouty, M. ; Aboulnasr, T. ; Bouchard, M.
Author_Institution :
Sch. of Inf. Technol. & Eng., Ottawa Univ., Ont.
Abstract :
A new algorithm for blind signal separation of speech signals that does not require pre-whitening is proposed in this paper. The algorithm is based on second order optimization using Riemannian geometry. The algorithm employs several practical approximations to the Hessian matrix of the maximum-likelihood blind separation cost function, to produce a computationally efficient algorithm that is capable of working on-line. Simulation results show the improved performance of the proposed algorithm with different mixing data
Keywords :
Hessian matrices; blind source separation; maximum likelihood estimation; speech processing; Riemannian geometry; blind signal separation; diagonal Hessian algorithm; maximum-likelihood blind separation cost function; second order optimization; speech signals; Blind source separation; Computational modeling; Convergence; Cost function; Geometry; Information technology; Information theory; Maximum likelihood estimation; Source separation; Speech;
Conference_Titel :
Acoustics, Speech and Signal Processing, 2006. ICASSP 2006 Proceedings. 2006 IEEE International Conference on
Conference_Location :
Toulouse
Print_ISBN :
1-4244-0469-X
DOI :
10.1109/ICASSP.2006.1661363
