Metraipolis › archive › dc_c480022bfb5d

Token Merging for Multilingual Speech Recognition: A Systematic Study Across Model Scale and Fine-Tuning — arXiv:2609.13151

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.13151
License (as archived)
CC0-1.0
Archived
2026-10-15T23:13:22.524Z
SHA-256 of the archived text
fc9bc3ca55b92c9b5ba6b6c76467aa971c0253db63893d90c5b2f64509e80e0e
Stable URL of this page
https://metraipolis.com/documento/dc_c480022bfb5d

How to cite

Token Merging for Multilingual Speech Recognition: A Systematic Study Across Model Scale and Fine-Tuning — arXiv:2609.13151 — arXiv (en), https://arxiv.org/abs/2609.13151 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c480022bfb5d (Metraipolis archive dc_c480022bfb5d, sha256 fc9bc3ca55b92c9b5ba6b6c76467aa971c0253db63893d90c5b2f64509e80e0e)
BibTeX
@misc{dcc480022bfb5d,
  title = {Token Merging for Multilingual Speech Recognition: A Systematic Study Across Model Scale and Fine-Tuning — arXiv:2609.13151},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.13151},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c480022bfb5d; sha256 fc9bc3ca55b92c9b5ba6b6c76467aa971c0253db63893d90c5b2f64509e80e0e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.13151. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Leading multilingual speech recognition models like Whisper transcribe diverse, low-resource languages without language-specific training but are computationally expensive to deploy. Token merging mitigates this inefficiency by dynamically combining redundant features, shortening the sequence length during inference without requiring retraining. In this paper, we systematically evaluate token merging on the Whisper model family across sixteen diverse languages and three different model sizes. We also test how token merging interacts with fine-tuning (DoRA) on low-resource languages. Our findings show that merging tokens increases computational efficiency with almost no loss in transcription accuracy across most low-resource languages and model sizes, and it works even after the model has been fine-tuned. Our results demonstrate that token merging is a highly practical method for making multilingual speech recognition faster and cheaper to deploy.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.13151 — 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 2026-10-15T22:51:09.808Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).