This paper presents CrossVoice, a novel cascade-based Speech-to-Speech
Translation (S2ST) system employing advanced ASR, MT, and TTS technologies with
cross-lingual prosody preservation through transfer learning. We conducted
comprehensive experiments comparing CrossVoice with direct-S2ST systems,
showing improved BLEU scores on tasks such as Fisher Es-En, VoxPopuli Fr-En and
prosody preservation on benchmark datasets CVSS-T and IndicTTS. With an average
mean opinion score of 3.75 out of 4, speech synthesized by CrossVoice closely
rivals human speech on the benchmark, highlighting the efficacy of
cascade-based systems and transfer learning in multilingual S2ST with prosody
transfer.

该论文介绍了 CrossVoice，这是一种采用先进的 ASR、MT 和 TTS 技术的级联式语音到语音翻译（S2ST）系统，通过迁移学习实现跨语言语调保留。我们通过与直接 S2ST 系统的全面实验比较，展示了在 Fisher Es-En、VoxPopuli Fr-En 等任务中，CrossVoice 在 BLEU 分数上的改进以及在基准数据集 CVSS-T 和 IndicTTS 上的语调保留能力。通过平均 4 分中的 3.75 的意见得分，CrossVoice 合成的语音与基准上的人类语音接近，突出了级联式系统和迁移学习在多语言 S2ST 中的有效性。