TY  - RPRT
TI  - VoxAudio: Vocalized Audio Synthesis via Multi-Reward Autoregressive Flow Matching
AU  - Wenxiang Guo
AU  - Changhao Pan
AU  - Ziyue Jiang
AU  - Zhou Zhao
AU  - Fei Wu
PY  - 2026
UR  - https://arxiv.org/abs/2608.12951
ID  - 2608.12951
ER  -