AI RESEARCH
Hear What Matters! Text-conditioned Selective Video-to-Audio Generation
arXiv CS.LG
•
ArXi:2512.02650v2 Announce Type: replace-cross