Metraipolis › archive › dc_65d13b9ef2d0

Dialect-Robust Speech Language Models with Synthetic Pseudo-Dialect Augmentation — arXiv:2610.09321

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.09321
License (as archived)
CC0-1.0
Archived
2026-10-08T11:26:01.331Z
SHA-256 of the archived text
fe1c1848355c270dfb2e9030f78fcce5500a4afa0cba12f3e4917f67df51bc5e
Stable URL of this page
https://metraipolis.com/documento/dc_65d13b9ef2d0

How to cite

Dialect-Robust Speech Language Models with Synthetic Pseudo-Dialect Augmentation — arXiv:2610.09321 — arXiv (en), https://arxiv.org/abs/2610.09321 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_65d13b9ef2d0 (Metraipolis archive dc_65d13b9ef2d0, sha256 fe1c1848355c270dfb2e9030f78fcce5500a4afa0cba12f3e4917f67df51bc5e)
BibTeX
@misc{dc65d13b9ef2d0,
  title = {Dialect-Robust Speech Language Models with Synthetic Pseudo-Dialect Augmentation — arXiv:2610.09321},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09321},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_65d13b9ef2d0; sha256 fe1c1848355c270dfb2e9030f78fcce5500a4afa0cba12f3e4917f67df51bc5e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09321. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Speech Language Model (SLM) performance often degrades on dialects due to data scarcity. Conventional text-to-speech (TTS) augmentation struggles to cover diverse dialects as it requires a certain amount of real dialect speech. We propose synthesizing pseudo-dialect speech by converting LLM-generated dialect text via a standard-language TTS model, requiring zero real dialect speech. Additionally, we introduce intermediate standard-text prediction during training, acting as semantic normalization for downstream tasks. We evaluate dialect understanding via dialect-to-English speech translation across Japanese, German, and Chinese dialects. Compared to synthetic standard speech baselines, pseudo-dialect augmentation improves scores for Japanese (from 25.38 to 26.24) and German (from 31.57 to 32.47). Furthermore, the intermediate standard-text prediction effectively bridges the semantic gap, boosting performance to 28.26 for Japanese and from 11.67 to 16.37 for Chinese. These results suggest that our approach scales to various languages without requiring speech resources specific to each dialect.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2610.09321 — 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 2026-10-08T10:24:06.313Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).