@misc{indiciae329327654cc8,
  title = {SARA: A Dual-Stream VAE for High-Fidelity Speech Generation via Integrating Semantic and Acoustic Representations},
  author = {Peijie Chen and Wenhao Guan and Weijie Wu and Kaidi Wang and Daiyu Huang and Zhuanling Zha and Junbo Li and Jun Fang and Qingyang Hong and Lin Li},
  year = {2026},
  url = {https://arxiv.org/abs/2606.11611},
  note = {Source identifier: 2606.11611}
}