Denoising Audio Signals in the Non-negative Auditory Cortical Domain

CÉSAR ERNESTO MARTÍNEZ; JOHN GODDARD CLOSE; LEANDRO EZEQUIEL DI PERSIA; DIEGO HUMBERTO MILONE; HUGO LEONARDO RUFINER

Córdoba

Congreso; Argentine Symposium on Technology; 2011

SADIO Sociedad Argentina de Informática.

In this work, a biologically-inspired denoising method foraudio signals is presented, which takes advantage of an approximationto the acoustical signal representation at the auditory cortical level. Itis based on an optimal dictionary of atoms, estimated from early auditoryspectrograms, and the Basis Pursuit algorithm to approximate thecortical activations. The proposed approach employs non-negative sparsecoding to pursue a simplified denoising algorithm which exploits a prioriinformation from both clean signals and noise. The method was appliedto artificial signals constructed from simultaneous chirps, corrupted withadditive noise. Results showed that using an objective quality measure,the method proposed here can improve the audio quality when it is appliedto noisy signals.