Abstract:

Recent research has shown remarkable performance in leveraging multiple extraneous conditional and non-mutually-exclusive seman- tic concepts for sound source separation, allowing the flexibility to extract a given target source based on multiple different queries.
In this work, we propose a new optimal condition training (OCT) method for single-channel target source separation, based on greedy parameter updates using the highest performing condition among equivalent conditions associated with a given target source. Our ex- periments show that the complementary information carried by the diverse semantic concepts significantly helps to disentangle and iso- late sources of interest much more efficiently compared to single- conditioned models. Moreover, we propose a variation of OCT with condition refinement, in which an initial condition vector is adapted to the given mixture and transformed to a more amenable repre- sentation for target source extraction. We showcase the effective- ness of OCT on diverse source separation experiments where it im- proves upon permutation invariant models with oracle assignment between estimated and target sources and obtains state-of-the-art performance in the more challenging task of text-based source sepa- ration, outperforming even dedicated text-only conditioned models.