Metraipolis › archive › dc_efcfb0cb9ceb

A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization — arXiv:2609.25471

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.25471
License (as archived): CC0-1.0
Archived: 2026-10-08T13:46:22.442Z
SHA-256 of the archived text: 6aea058085f2d2258e36a3ada88e7146ad217695691374c6209a1061b9b36f61
Stable URL of this page: https://metraipolis.com/documento/dc_efcfb0cb9ceb

How to cite

A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization — arXiv:2609.25471 — arXiv (en), https://arxiv.org/abs/2609.25471 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_efcfb0cb9ceb (Metraipolis archive dc_efcfb0cb9ceb, sha256 6aea058085f2d2258e36a3ada88e7146ad217695691374c6209a1061b9b36f61)

BibTeX

@misc{dcefcfb0cb9ceb,
  title = {A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization — arXiv:2609.25471},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.25471},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_efcfb0cb9ceb; sha256 6aea058085f2d2258e36a3ada88e7146ad217695691374c6209a1061b9b36f61},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.25471. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Semi-supervised federated learning (SSFL) trains models on clients' unlabeled data using a teacher to generate pseudo-labels, with a small labeled seed dataset on the server. Automatic Speech Recognition (ASR) is particularly fragile here: pseudo-label errors compound across the output sequence and across training rounds into divergence, leaving a large gap to fully-supervised FL. We show that closing this gap turns on two coupled design axes -- the teacher (which model generates the pseudo-labels) and the anchor (the server-side updates on labeled data that stabilize training). On the teacher axis, a per-client online teacher (each client's own evolving model) diverges on its own, but once stabilized it matches or beats the broadcast global teacher (one server model, fixed within a round) -- decisively in-domain and competitively under domain shift. As the seed grows stronger and the online teacher's advantage narrows, a transitioning teacher (global $\rightarrow$ online at round $r$) matches or beats both. On the anchor axis, the server must keep training on labeled data between rounds -- otherwise the online teacher drifts -- and this interleaving, more than the seed model, governs convergence. The two axes are inseparable: aggressive teacher choices pay off only once the anchor stabilizes training, which is highly sensitive to data augmentation and batch size -- the settings that govern how much input and gradient noise the server injects. How much stabilization is needed is domain-dependent, governed by the dispersion of the seed data and its overlap with client data. These findings yield guidelines for SSFL in ASR training, improving over the strongest prior method on 9 of 11 pairs, by $20.8\%$ on average in-domain and $10.0\%$ cross-domain, narrowing the gap to fully-supervised FL.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.AI).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2609.25471 — 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 2026-10-08T13:18:38.230Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).