Metraipolis › archive › dc_4abc608a4752

Tacit-TTS: From Autoregressive Decoding to Masked Prediction for Efficient Transcript-Free Voice Cloning — arXiv:2609.38658

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.38658
License (as archived)
CC0-1.0
Archived
2026-10-07T13:14:20.076Z
SHA-256 of the archived text
6c06c0ff50ae0bd17890a7eabb48b853086727c802161197482994b774dd26c0
Stable URL of this page
https://metraipolis.com/documento/dc_4abc608a4752

How to cite

Tacit-TTS: From Autoregressive Decoding to Masked Prediction for Efficient Transcript-Free Voice Cloning — arXiv:2609.38658 — arXiv (en), https://arxiv.org/abs/2609.38658 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4abc608a4752 (Metraipolis archive dc_4abc608a4752, sha256 6c06c0ff50ae0bd17890a7eabb48b853086727c802161197482994b774dd26c0)
BibTeX
@misc{dc4abc608a4752,
  title = {Tacit-TTS: From Autoregressive Decoding to Masked Prediction for Efficient Transcript-Free Voice Cloning — arXiv:2609.38658},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.38658},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4abc608a4752; sha256 6c06c0ff50ae0bd17890a7eabb48b853086727c802161197482994b774dd26c0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38658. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

TTS systems with autoregressive semantic modeling have demonstrated strong zero-shot voice cloning performance and rich expressive variation, but their sequential decoding incurs substantial latency. Non-autoregressive alternatives offer much faster generation, yet often rely on more restrictive reference conditioning, such as requiring transcripts of the reference speech during inference. We present Tacit-TTS, an efficient transcript-free zero-shot voice cloning system distilled from IndexTTS2. Our model replaces autoregressive text-to-semantic decoding with masked non-autoregressive generation, introduces training-free acoustic length estimation, and accelerates the flow-matching renderer through ReFlow distillation. Across two English and two Mandarin datasets, Tacit-TTS achieves competitive zero-shot quality while generating speech over 10x faster than IndexTTS2 for utterances longer than 5 seconds. Its transcript-free conditioning further supports cross-lingual and non-lexical references. We validate this capability using references from eight other languages, infant babble, and synthetic gibberish, where transcript-dependent systems often degrade or fail due to unreliable ASR transcripts.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CL).
Categorie arXiv: Audio and Speech Processing (eess.AS) ; Computation and Language (cs.CL); Sound (cs.SD).
Pagina del lavoro: https://arxiv.org/abs/2609.38658 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T12:44:51.939Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).