ISCA Tutorial and Research Workshop on Statistical and Perceptual Audition (SAPA2008)

Brisbane, Australia
September 21, 2008

Computational Auditory Induction by Missing-Data Non-Negative Matrix Factorization

Jonathan Le Roux (1,3), Hirokazu Kameoka (2), Nobutaka Ono (1), Alain de Cheveigné (3), Shigeki Sagayama (1)

(1) Graduate School of Information Science and Technology, The University of Tokyo, Japan
(2) NTT Communication Science Laboratories, NTT Corporation, Japan
(3) CNRS, Université Paris 5, and Ecole Normale Supérieure, France

The human auditory system has the ability, known as auditory induction, to estimate the missing parts of a continuous auditory stream briefly covered by noise and perceptually resynthesize them. Humans are thus able to simultaneously analyze an auditory scene and reconstruct the underlying signal. In this article, we formulate this ability as a non-negative matrix factorization (NMF) problem with unobserved data, and show how to solve it using an auxiliary function method. We explain how this method can also be generally related to the EM algorithm, enabling the use of prior distributions on the parameters. We show how sparseness is a key to global feature extraction, and that our method is ideally able to extract patterns which never occur completely. We finally illustrate on an example how our method is able to simultaneously analyze a scene and interpolate the gaps into it.

Full Paper

Bibliographic reference.  Roux, Jonathan Le / Kameoka, Hirokazu / Ono, Nobutaka / Cheveigné, Alain de / Sagayama, Shigeki (2008): "Computational auditory induction by missing-data non-negative matrix factorization", In SAPA-2008, 1-6.