Metraipolis › archive › dc_8092989e7dc4

Choosing a PEFT Variant for Per-Patient Dysarthric ASR: A Single-Speaker Case Study on Two ASR Bases — arXiv:2609.02735

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.02735
License (as archived)
CC0-1.0
Archived
2026-10-17T04:16:56.241Z
SHA-256 of the archived text
2ddb50ed8bdad5ad95249edd9b546517a0aad96554290a6e7b5aeff5b252556f
Stable URL of this page
https://metraipolis.com/documento/dc_8092989e7dc4

How to cite

Choosing a PEFT Variant for Per-Patient Dysarthric ASR: A Single-Speaker Case Study on Two ASR Bases — arXiv:2609.02735 — arXiv (en), https://arxiv.org/abs/2609.02735 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8092989e7dc4 (Metraipolis archive dc_8092989e7dc4, sha256 2ddb50ed8bdad5ad95249edd9b546517a0aad96554290a6e7b5aeff5b252556f)
BibTeX
@misc{dc8092989e7dc4,
  title = {Choosing a PEFT Variant for Per-Patient Dysarthric ASR: A Single-Speaker Case Study on Two ASR Bases — arXiv:2609.02735},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.02735},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8092989e7dc4; sha256 2ddb50ed8bdad5ad95249edd9b546517a0aad96554290a6e7b5aeff5b252556f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02735. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Per-patient adapters are the preferred production architecture for dysarthric automatic speech recognition (ASR), yet parameter-efficient fine-tuning (PEFT) variants have not been compared in the speaker-dependent, per-patient regime. We present a single-speaker case study comparing seven LoRA-family methods (LoRA, QLoRA, AdaLoRA, DoRA, LoHA, VeRA, VB-LoRA) on two production bases (Whisper-large-v3 with Hungarian fine-tuning, and a multilingual Qwen3-ASR-1.7B checkpoint) for one post-stroke Hungarian male speaker (S1, 409 utterances; severe dysarthria on auditory-perceptual clinical assessment). Attention-projection adapters substantially improve CER on both bases. Across three seeds, a paired bootstrap detects no significant LoRA-DoRA difference (p>0.5; 13.86/13.90 % CER on Whisper, 28.10/28.33 % on Qwen3-ASR), so we adopt the simpler, cheaper LoRA. Real 4-bit (NF4) QLoRA is worse on every seed and both bases (14.56/30.09 % CER) with no memory saving at this scale, and LoHA, VeRA, VB-LoRA and AdaLoRA do not reach the LoRA family, though LoHA still gives an 18.6 % relative CER reduction on Whisper. On the same base, full fine-tuning is more accurate (11.43 % CER), but a 115 MB LoRA that also adapts the feed-forward blocks reaches within 0.66 pp of it at approximately 3.7 % of the per-patient storage. A 6-point enrollment grid shows about 5 min of patient audio captures 45.6 % of the zero-shot-to-30-min CER reduction, with further gains at 10 and 30 min (caveat: one speaker, one language, severe post-stroke dysarthria). Training scripts and recipes will be released, source-available under a research-use licence, on publication.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Sound (cs.SD).
Pagina del lavoro: https://arxiv.org/abs/2609.02735 — 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 2026-10-17T03:26:49.810Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).