Probabilistic Modelling of Signal Mixtures with Differentiable Dictionaries
This work addresses the challenge of signal mixture modeling for audio processing, but appears incremental as it builds on existing non-negative matrix factorization techniques.
The authors tackled the problem of modeling signal mixtures with non-linear sources by introducing differentiable dictionary search, a method for incorporating prior information into non-negative matrix factorization, and demonstrated its effectiveness on an audio decomposition task with controlled studies.
We introduce a novel way to incorporate prior information into (semi-) supervised non-negative matrix factorization, which we call differentiable dictionary search. It enables general, highly flexible and principled modelling of mixtures where non-linear sources are linearly mixed. We study its behavior on an audio decomposition task, and conduct an extensive, highly controlled study of its modelling capabilities.