1 paper · 1 filter
Chenxu Xiong, Ruibo Fu, Shuchen Shi +9
Current mainstream audio generation methods primarily rely on simple text prompts, often failing to capture the nuanced details necessary for multi-style audio generation. To addre…