I'm interested in generative models.
SSV2A. Gotta Hear Them All: Towards Sound Source Aware Audio Generation.
CoP. The Collapse of Patches