Metraipolis › archive › dc_cc034408d8e8
ParA-LLM: A Unified Approach to Paralinguistic and Acoustic Speech Understanding — arXiv:2609.22771
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.22771
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T07:08:17.434Z
- SHA-256 of the archived text
96ed4c888bbe5b0df33fbb0186796e3dd976cb0cbaf2e6f3a0dff1abe813e9e8- Stable URL of this page
https://metraipolis.com/documento/dc_cc034408d8e8
How to cite
ParA-LLM: A Unified Approach to Paralinguistic and Acoustic Speech Understanding — arXiv:2609.22771 — arXiv (en), https://arxiv.org/abs/2609.22771 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cc034408d8e8 (Metraipolis archive dc_cc034408d8e8, sha256 96ed4c888bbe5b0df33fbb0186796e3dd976cb0cbaf2e6f3a0dff1abe813e9e8)
BibTeX
@misc{dccc034408d8e8,
title = {ParA-LLM: A Unified Approach to Paralinguistic and Acoustic Speech Understanding — arXiv:2609.22771},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.22771},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cc034408d8e8; sha256 96ed4c888bbe5b0df33fbb0186796e3dd976cb0cbaf2e6f3a0dff1abe813e9e8},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.22771. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recent advances in Audio LLMs have achieved human-level speech recognition, yet existing systems struggle to capture paralinguistic aspects such as speaker traits, expressive variations, and environmental acoustic conditions. To address this, we design a framework of 22 paralinguistic characteristics and create a dataset of over 1.2M Audio-QA pairs. We develop ParA-LLM, trained with a two-stage curriculum: first on single-attribute questions to build foundational knowledge, then on multi-attribute questions for joint reasoning over speaker and acoustic characteristics. We also release ParA-Bench, a benchmark of 6,000 multiple-choice questions across speaker-speech, acoustic, and mixed categories, where frontier models like GPT-4o-Audio achieve only 36% accuracy. ParA-LLM surpasses state-of-the-art Audio LLMs like GPT-4o-Audio by 7.5% on ParA-Bench, with additional gains of 1.13% on MMAU-Pro Speech and 7.49% on MMAR Speech. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-22 in cs.LG). Categorie arXiv: Sound (cs.SD) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS); Signal Processing (eess.SP). Pagina del lavoro: https://arxiv.org/abs/2609.22771 — 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 2026-10-09T06:58:09.530Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).