English

mHuBERT-147: A Compact Multilingual HuBERT Model

Computation and Language 2024-11-22 v5 Sound Audio and Speech Processing

Abstract

We present mHuBERT-147, the first general-purpose massively multilingual HuBERT speech representation model trained on 90K hours of clean, open-license data. To scale up the multi-iteration HuBERT approach, we use faiss-based clustering, achieving 5.2x faster label assignment than the original method. We also apply a new multilingual batching up-sampling strategy, leveraging both language and dataset diversity. After 3 training iterations, our compact 95M parameter mHuBERT-147 outperforms larger models trained on substantially more data. We rank second and first on the ML-SUPERB 10min and 1h leaderboards, with SOTA scores for 3 tasks. Across ASR/LID tasks, our model consistently surpasses XLS-R (300M params; 436K hours) and demonstrates strong competitiveness against the much larger MMS (1B params; 491K hours). Our findings indicate that mHuBERT-147 is a promising model for multilingual speech tasks, offering an unprecedented balance between high performance and parameter efficiency.

Keywords

Cite

@article{arxiv.2406.06371,
  title  = {mHuBERT-147: A Compact Multilingual HuBERT Model},
  author = {Marcely Zanon Boito and Vivek Iyer and Nikolaos Lagos and Laurent Besacier and Ioan Calapodescu},
  journal= {arXiv preprint arXiv:2406.06371},
  year   = {2024}
}

Comments

Extended version of the Interspeech 2024 paper of same name

R2 v1 2026-06-28T16:59:47.071Z