Metraipolis › archive › dc_806d39e104e2
Conversational Voice Aesthetic Model with Reinforcement Learning from Human Listeners — arXiv:2610.10868
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.10868
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T06:13:38.070Z
- SHA-256 of the archived text
960fa63c4dc6992b6b6ffc5cb11cd5dce657f85a8d9c655673f21d55a459df5c- Stable URL of this page
https://metraipolis.com/documento/dc_806d39e104e2
How to cite
Conversational Voice Aesthetic Model with Reinforcement Learning from Human Listeners — arXiv:2610.10868 — arXiv (en), https://arxiv.org/abs/2610.10868 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_806d39e104e2 (Metraipolis archive dc_806d39e104e2, sha256 960fa63c4dc6992b6b6ffc5cb11cd5dce657f85a8d9c655673f21d55a459df5c)
BibTeX
@misc{dc806d39e104e2,
title = {Conversational Voice Aesthetic Model with Reinforcement Learning from Human Listeners — arXiv:2610.10868},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.10868},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_806d39e104e2; sha256 960fa63c4dc6992b6b6ffc5cb11cd5dce657f85a8d9c655673f21d55a459df5c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.10868. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We introduce Conversational Voice Aesthetic Model, a speech large language model for describing the voice aesthetics of real or synthetic speech responses in natural conversational contexts. Given a context and a response speech, CVAM describes salient moments that characterize the voice and predicts nine categorical attributes spanning gender, pitch, pacing, emotion, and delivery. The key challenge lies in perceptual fields such as emotion and delivery, which are inherently subjective and lack definitive ground truth. Therefore, we collect ~10 human annotations for each of 3k real and synthetic responses derived from the CANDOR corpus. CVAM is supervised finetuned on synthesized aesthetic descriptions and labels, then optimized with Group Relative Policy Optimization on human judgments. Experiments show that CVAM better agrees with human listeners than Gemini 3.1 Pro and open-source speech LLMs, and outperforms single-human-vs.-rest agreement. Together, we demonstrate the importance of grounding voice aesthetics in human perception and propose a principled framework for human alignment. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Audio and Speech Processing (eess.AS) ; Computation and Language (cs.CL); Machine Learning (cs.LG); Multimedia (cs.MM); Sound (cs.SD). Pagina del lavoro: https://arxiv.org/abs/2610.10868 — 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 2026-10-09T05:47:15.411Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).