SonicWorld: Source-Centric Latent Dynamics for Spatial Audio Reasoning
Dynamic spatial audio reasoning requires a model to associate changing acoustic observations with the correct source and recover its spatial state at a queried time or interval, a challenge intensified by overlapping speech and source motion. To study this problem, we introduce Sonic-Bench, comprising 2.8K first-order...
Chen-Long Wei, Han-Ling Wang, Mingjin Lou et al.
· 0 citations