On information maximization and blind signal deconvolution
Résumé
In the following paper we investigate a recent algorithm for blind signal deconvolution and show that the algorithm is appropriate to solve the deconvolution problem only, if the deconvolution filter is constrained to be minimum phase. We improve the algorithm such that this constraint is removed and present experimental results that demonstrate the improved properties of the extended algorithm. Moreover, the experimental results show that the fixed signal density model implemented in the original algorithm has to be extended also, to achieve a proper objective function for the general deconvolution problem.
Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...