Non-negative Hidden Markov Modeling of Audio with Application to Source Separation

LVA/ICA - International Conference on Latent Variable Analysis and Signal Separation, September 2010

Published September 27, 2010

Gautham Mysore, Paris Smaragdis, B. Raj

Best Student Paper Award

In recent years, there has been a great deal of work in modeling audio using non-negative matrix factorization and its probabilistic counterparts as they yield rich models that are very useful for source separation and automatic music transcription. Given a sound source, these algorithms learn a dictionary of spectral vectors to best explain it. This dictionary is however learned in a manner that disregards a very important aspect of sound, its temporal structure. We propose a novel algorithm, the non-negative hidden Markov model (N-HMM), that extends the aforementioned models by jointly learning several small spectral dictionaries as well as a Markov chain that describes the structure of changes between these dictionaries. We also extend this algorithm to the non-negative factorial hidden Markov model (N-FHMM) to model sound mixtures, and demonstrate that it yields superior performance in single channel source separation tasks.

Learn More

Research Areas:  AI & Machine Learning Audio