AfriHuBERT: A self-supervised speech representation model for African languages

README

This is the pre-trained model for paper

AfriHuBERT: A self-supervised speech representation model for African languages, accepted by Interspeech 2025

By Jesujoba O. Alabi, Xuechen Liu, Dietrich Klakow, Junichi Yamagishi

The code for building and using the pre-trained model is on github.

Please cite the paper if you use the pre-trained models in your work.

@inproceedings{alabi2024afrihubert,
      title={AfriHuBERT: A self-supervised speech representation model for African languages}, 
      author={Jesujoba O. Alabi and Xuechen Liu and Dietrich Klakow and Junichi Yamagishi},
      year={2025},
      booktitle={Proc. Interspeech} 
}

COPYING

This pretrained model is licensed under the Creative Commons Attribution Non Commercial Share Alike 4.0 International  http://creativecommons.org/licenses/by/4.0/legalcode. Please see `LICENSE.txt` for the terms and conditions of this pretrained model.

ACKNOWLEDGMENTS

This work was conducted during the first author’s internship at NII, Japan. This study is partially supported by JST AIP Acceleration Research (JPMJCR24U3). Part of this study was carried out using the TSUBAME4.0 supercomputer at the Institute of Science Tokyo. Also, we thank Xin Wang, Badr M. Abdullah, Siyang Wang, Wanying Ge, David Adelani, and Aravind Krishnan for their helpful feedback.