PhD student, Seoul National University
1 paper at NeurIPS 2025
We propose unified latent diffusion framework for simultaneous music generation, source imputation, and query-driven arbitrary source extraction.
