A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization
arXiv:2609.25471v1 Announce Type: new Abstract: 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-labe…
Read the full story at arXiv cs.LG ↗
Timeline · 1 report
- 2026-09-23 04:00 · arXiv cs.LG
A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization