aranemini/northern-kurdish-pseudolabel

Northern Kurdish Raw Audio Collection

Dataset Summary

This repository contains a large collection of raw Northern Kurdish (Kurmanji Kurdish) speech recordings gathered from publicly available Kurdish media sources.

The corpus was assembled to support research and development in:

• Automatic Speech Recognition (ASR)
• Speech Translation (ST)
• Text-to-Speech (TTS)
• Self-Supervised Learning (SSL)
• Spoken Language Understanding (SLU)
• Low-Resource Speech Processing

The recordings were collected primarily from Kurdish media outlets, with the majority originating from Sterk TV, and include news programs, interviews, discussions, cultural programs, documentaries, and other forms of spontaneous speech.

Unlike curated audiobook corpora, this collection mainly contains naturally occurring broadcast speech and conversational content, providing a valuable resource for building robust speech technologies for Kurdish.

Dataset Statistics

Property	Value
Language	Northern Kurdish (Kurmanji Kurdish, kmr)
Speech duration	~2,000 hours
Number of audio files	~760,000
Annotation	Pseudo-labeled transcriptions
Domain	Broadcast media, interviews, news, spontaneous speech

Disclaimer

The audio recordings included in this repository were collected from publicly accessible Kurdish media sources for research purposes.

If you are the copyright holder, author, narrator, publisher, broadcaster, or rights owner of any material included in this dataset and do not wish your content to be distributed through this repository, please contact:

emini.aran@gmail.com

Upon verification of the request, the corresponding content will be removed from future releases of the dataset.

Citation

If you use this dataset, please cite:

@inproceedings{mohammadamini25_interspeech,
  title     = {Scaling pseudo-labeling data for end-to-end low-resource speech translation (the case of Kurdish language)},
  author    = {Mohammad Mohammadamini and Aghilas Sini and Marie Tahon and Antoine Laurent},
  booktitle = {Interspeech 2025},
  year      = {2025},
  pages     = {898--902},
  doi       = {10.21437/Interspeech.2025-887}
}

@inproceedings{mohammadamini2025frugal,
  title={Apprentissage de modèles frugaux pour les langues peu dotées à partir de larges modèles d'ASR},
  author={Mohammadamini, Mohammad},
  school={Le Mans Université},
  year={2025}
}