Zurück zur Liste

Perceptual Sparsity Modeled by Simultaneous Masking

ID 6194
Abstract Human auditory perception can be described as sparse, as not all components in an audio signal are perceived. We present an efficient algorithm for determining and removing perceptually irrelevant time-frequency components of a sound. The algorithm is based on a model of simultaneous masking in the auditory system. It computes a sparse time-frequency representation by detecting perceptually irrelevant components. A ma jor goal was its applicability to any complex sound, and therefore coping with the absence of a clear separation between maskers and targets, as well as its computational efficiency. The algorithm determines the masked threshold for each spectral component within an analysis window. The masked threshold function is then shifted in level by an amount determined experimentally, and all components falling below the shifted function (the irrelevance threshold) are removed. Thirty-six normal hearing sub jects participated in an experiment to determine the maximum shift value for which they could not discriminate the irrelevance filtered signal from the original signal. On average across the test stimuli, 36 percent of the time-frequency components fell below the irrelevance threshold. The mathematical foundations of the applied analysis-resynthesis system, a discrete Gabor system, is a time-variant filtering process called Gabor filter.
Edition TELECOM ParisTech Workshop on Music Signal Processing in Paris (Frankreich)
Monat 07
Name CNRS (Frankreich)
Ort CNRS (Frankreich)
Publikationsart Poster
Jahr 2008
AutorInnen Balasz, P., Laback, B., Eckel, G., Deutsch, W.