Metraipolis › archive › dc_fd5ec4c7d87a
Preference Optimization for Non-Verbal Vocalization Synthesis — arXiv:2608.24163
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.24163
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T06:36:43.525Z
- SHA-256 of the archived text
089227fcc5105df08663d1e6eef2cbfc69ac5d544f1b80afdb62a30a5785566e- Stable URL of this page
https://metraipolis.com/documento/dc_fd5ec4c7d87a
How to cite
Preference Optimization for Non-Verbal Vocalization Synthesis — arXiv:2608.24163 — arXiv (en), https://arxiv.org/abs/2608.24163 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fd5ec4c7d87a (Metraipolis archive dc_fd5ec4c7d87a, sha256 089227fcc5105df08663d1e6eef2cbfc69ac5d544f1b80afdb62a30a5785566e)
BibTeX
@misc{dcfd5ec4c7d87a,
title = {Preference Optimization for Non-Verbal Vocalization Synthesis — arXiv:2608.24163},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.24163},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fd5ec4c7d87a; sha256 089227fcc5105df08663d1e6eef2cbfc69ac5d544f1b80afdb62a30a5785566e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.24163. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Non-verbal vocalizations (NVs), such as laughter, coughs, and sighs, are essential for expressive TTS, but the effectiveness of preference optimization for NV generation remains poorly understood. We systematically study preference optimization for NV-capable TTS, focusing on preference signals, preference-pair construction, and DPO-based optimization objectives. We formulate an NV-aware character error rate (NV-CER) by treating NV tags as distinct output symbols and computing a weighted pinyin-based CER over both verbal and non-verbal content, enabling controllable optimization of NV realization without modifying the underlying optimization algorithm. Experiments on Emilia-NV and the augmented NV-Bench covering 18 NV types reveal how different design choices affect NV realization and lexical fidelity, and establish an effective setup using standard DPO. Objective, LLM-based, and human evaluations provide converging evidence for our findings, offering practical insights into NV-aware post-training for expressive TTS. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-26 in cs.LG). Categorie arXiv: Audio and Speech Processing (eess.AS) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2608.24163 — 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 2026-10-18T06:28:52.341Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-26?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).