TY  - RPRT
TI  - Agentic-DuplexGen: Decoupling Content, Timing, and Acoustics for Synthetic Dialogue Speech
AU  - Pengcheng Wang
AU  - Sheng Li
AU  - Jiyi Li
AU  - Takahiro Shinozaki
PY  - 2026
UR  - https://arxiv.org/abs/2608.16053
ID  - 2608.16053
ER  -