NMF with time-frequency activations to model non stationary audio events

Real world sounds often exhibit non-stationary spectral characteristics such as those produced by a harpsichord or a guitar. The classical Non-negative Matrix Factorization (NMF) needs a number of atoms to accurately decompose the spectrogram of such sounds. An extension of NMF is proposed hereafter which includes timefrequency activations based on ARMA modeling. This leads to an efficient single-atom decomposition for a single audio event. The new algorithm is tested on real audio data and shows promising results.

• Type: Conference
• Year: 2010
• Where: ICASSP
• Authors: Romain Hennequin, Roland Badeau, Bertrand David
• Keywords: Efficient Single-atom Decomposition, ICASSP 2010, Non-Negative Matrix Factorization, Non-stationary Spectral Characteristics, Signal Processing