Random models for audio signals expansion on hybrid MDCT dictionaries

Matthieu Kowalski (1) , Bruno Torrésani (1)

1 LATP - Laboratoire d'Analyse, Topologie, Probabilités

A new approach for signal expansion with respect to hybrid dictionaries, based upon probabilistic modeling is proposed and studied, with emphasis on audio signal processing applications. The signal is modeled as a sparse linear combination of waveforms, taken from the union of two orthonormal bases, with random coefficients. The behavior of the analysis coefficients, namely inner products of the signal with all basis functions, is studied in details, which shows that these coefficients may generally be classified in two categories: significant coefficients versus unsignificant coefficients. Conditions ensuring the feasibility of such a classification are given. When the classification is possible, it leads to efficient estimation algorithms, that may in turn be used for de-noising or coding purpose. The proposed approach is illustrated by numerical experiments on audio signals, using MDCT bases.

Mots clés: Sparse Representations; Non-linear signal approximation; Time-frequency decompositions; Denoising.; Denoising

Domaines: Traitement du signal et de l'image [eess.SP]; Probabilités [math.PR]; Traitement du signal et de l'image [eess.SP]

Matthieu Kowalski, Bruno Torrésani. Random models for audio signals expansion on hybrid MDCT dictionaries. 2007. ⟨hal-00142088v1⟩