1 paper
Yufan Wen, Zhaocheng Liu, YeGuo Hua +4
Synthesizing coherent soundtracks for long-form videos remains a formidable challenge, currently stalled by three critical impediments: computational scalability, temporal coherenc…