CrossVoice: Crosslingual Prosody Preserving Cascade-S2ST using Transfer Learning
This addresses the problem of maintaining natural prosody in multilingual speech translation for users of cascade-based systems, though it is incremental as it builds on existing ASR, MT, and TTS technologies.
The paper tackled speech-to-speech translation with cross-lingual prosody preservation, achieving improved BLEU scores on tasks like Fisher Es-En and VoxPopuli Fr-En, and a mean opinion score of 3.75 out of 4 on benchmark datasets.
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.