Rotational reset strategy for online semi-supervised NMF-based speech enhancement for long recordings

Non-negative matrix factorization (NMF) has been successfully applied to speech enhancement in non-stationary noisy environments. Recently proposed online semi-supervised NMF algorithms are of particular interest as they carry the two nice properties (online and semi-supervised) of classical speech enhancement approaches. These algorithms, however, have only been evaluated using noisy mixtures shorter than 30 seconds. In this paper we find that these algorithms work well when it is run for less than 1 minute, but degradation of the enhanced speech signal starts to appear after 2 minutes. We analyze that the reason is due to the inappropriate dictionary update rule, which gradually loses its ability in updating the speech dictionary. We then propose a simple rotational reset strategy to solve the problem: Instead of continuously updating the entire speech dictionary, we periodically and rotationally select elements and reset their values to random numbers. Experiments show that this strategy successfully solves the degradation problem and the improved algorithm outperforms classical speech enhancement algorithms significantly even when they are run for 10 minutes.