A Recurrent Neural Filter for Adaptive Noise Cancellation
Δεν υπάρχει διαθέσιμη μικρογραφία
Ημερομηνία
2006
Τίτλος Εφημερίδας
Περιοδικό ISSN
Τίτλος τόμου
Εκδότης
Δικαιώματα
Attribution-NonCommercial-NoDerivatives 4.0 Διεθνές
Άδειες
Παραπομπή
Παραπομπή
Περίληψη
This paper presents a dynamic neural filter for adaptive
noise cancellation. The cancellation task is transformed to
a system-identification problem, which is tackled by use
of the Block-Diagonal Recurrent Neural Network. The
filter is applied to a benchmark noise cancellation
problem, where a comparative analysis with a series of
other dynamic models is conducted, underlining the
effectiveness of the proposed filter and its superior
performance over its competing rivals.