Metraipolis › archive › dc_a94f04faac8e

Anuvani-LID: Robust and Calibrated Spoken Language Identification for Indian Languages — arXiv:2610.20263

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.20263
License (as archived)
CC0-1.0
Archived
2026-10-15T20:35:39.339Z
SHA-256 of the archived text
4a032e7541ddc31224c06ee10724a54b9f2c0d0b851afaee5905598e4620ae02
Stable URL of this page
https://metraipolis.com/documento/dc_a94f04faac8e

How to cite

Anuvani-LID: Robust and Calibrated Spoken Language Identification for Indian Languages — arXiv:2610.20263 — arXiv (en), https://arxiv.org/abs/2610.20263 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a94f04faac8e (Metraipolis archive dc_a94f04faac8e, sha256 4a032e7541ddc31224c06ee10724a54b9f2c0d0b851afaee5905598e4620ae02)
BibTeX
@misc{dca94f04faac8e,
  title = {Anuvani-LID: Robust and Calibrated Spoken Language Identification for Indian Languages — arXiv:2610.20263},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.20263},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a94f04faac8e; sha256 4a032e7541ddc31224c06ee10724a54b9f2c0d0b851afaee5905598e4620ae02},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.20263. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Every multilingual speech pipeline begins with a language decision, an error there propagates silently through recognition, translation and routing. In India it spans 22 scheduled languages across four language families, heard over short, noisy, narrowband and code-mixed audio. Foundational speech models increasingly present language identification (LID) as an inherent latent, generalized capability. Autoregressive decoders in speech models emit a special language ID token, and modern LLMs with audio capabilities can name the language when prompted. We test this assumption on identical clips and find that whisper and similar speech models has tokens for only 15 of the 23 languages, on one second inputs Gemini 3.5 Flash reaches 0.52 macro-F1 and GPT-audio 0.14. We present Anuvani LID, a dedicated 115M parameter LID conformer model trained on 54,643 hours of Indian speech spanning 22 Indian languages and Indian accented English. Along with designed and controlled studies on 102 condition benchmark with speaker and corpus held out tests. Across 37 conditions, it reaches 0.914 mean macro-F1, 5.2 points above Gemini 3.5 Flash and ahead of 14 other systems. It is 24 points ahead of Gemini at one second, scores 0.897 on all 23 languages where no other system exceeds 0.713, and is calibrated to an ECE of 0.008. Reliable Indian LID still requires a robust, dedicated and calibrated model.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.20263 — 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 2026-10-15T19:36:18.249Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).