Kaidi Wang Yi He Wenhao Guan Weijie Wu Hongwu Ding Xiong Zhang Di Wu Meng Meng Jian Luan 0001 Lin Li Qingyang Hong SyncVoice: Towards Video Dubbing with Vision-Augmented Pretrained TTS Model. 2025 December abs/2512.05126 CoRR https://doi.org/10.48550/arXiv.2512.05126 db/journals/corr/corr2512.html#abs-2512-05126 streams/journals/corr