Sinusoidal modeling using psychoacoustic-adaptive matching pursuits

Abstract

Abstract-In this letter, we propose a segment-based matchingpursuit algorithm where the psychoacoustical properties of the human auditory system are taken into account. Rather than scaling the dictionary elements according to auditory perception, we define a psychoacoustic-adaptive norm on the signal space that can be used for assigning the dictionary elements to the individual segments in a rate-distortion optimal way. The new algorithm is asymptotically equal to signal-to-mask-ratio-based algorithms in the limit of infinite-analysis window length. However, the new algorithm provides a significantly improved selection of the dictionary elements for finite window length.

On perceptual distortion minimization and nonlinear least-squares frequency estimation

“…A recently published psychoacoustic masking model for audio coding has been shown to form a distortion measure [30], [31], and this distortion measure has been applied successfully to the sinusoidal estimation problem in [15], [23], [32], [33]. Based on this we define the perceptual frequency estimation problem and its optimal solution.…”

Sparse denoising of audio by greedy time-frequency shrinkage

“…Despite their simplicity, greedy algorithms remain a popular choice in image and audio processing application due to their good performance, computational efficiency and ability to produce compact signal representations. As a result, in recent years several low bit-rate audio coding and compression approaches based on greedy atomic decomposition have been introduced [20,21,22,23]. One of the most significant aspects of greedy signal representation algorithms is their ability to work with redundant dictionaries.…”

“…As a result, several variants of the MP algorithm have been proposed, each usually tailored to a specific application. We do not attempt to list or review all of these MP variants here, rather we refer the reader to the literature [38,22,23,39,32]. Some of these extensions include dictionary weighting, integrating psychoacoustic information and improving computational efficiency.…”

Analysis, Visualization, and Transformation of Audio Signals Using Dictionary-based Methods†

“…Molecular Matching Pursuit (MMP) [Daudet, 2006, Leveau et al, 2008 takes advantage of knowing the structures expected in acoustic signals, i.e., transients and tonals, to create sparse and structured models of each. Psychoacoustic Adaptive MP [Heusdens et al, 2002] provides a perceptual measure for selecting the atoms of a model. And Cyclic Matching Pursuit (CMP) [Christensen and Jensen, 2007] refines the estimates of the parameters of atoms in the model to reduce its distortion.…”