anyantudre/MooreSpeechCorpora

Moore Speech Corpora: A Cleaned, Denoised Audio-Text Dataset for Mooré TTS and ASR

The Moore Speech Corpora is a collection of aligned audio and text in Mooré, gathered from publicly available sources. This unified corpus is curated for research and academic purposes in low-resource speech and language processing, especially for text-to-speech (TTS) and automatic speech recognition (ASR) in the Mooré language (ISO 639-3: mos).

Mooré is under-represented in current speech corpora and technologies. Our mission is to make AI accessible to all communities. By open-sourcing this dataset and its associated tools, we aim to promote and preserve our national languages through practical speech-technology resources.

Currently this version of the dataset is a merged, denoised, and ready-to-use audio-text corpus built from:

• Moore Speech Contes
• Moore Speech Devinettes
• Moore Speech Proverbs

We're still working to add other sources.

⚠️ Access is gated. To request access, please read the policy below.
🚩 TLDR: Anonymous or vague requests will not be approved.

All scripts used to prepare this dataset are part of the Moore Speech Corpora Toolkit. ⭐ if it’s helpful!

---

Dataset Structure

Each record contains:

• audio: Original 16 kHz WAV audio (raw)
• denoised_audio: Denoised 16 kHz WAV audio (cleaned from background music)
• text: Transcription in Mooré
• duration: Duration in seconds
• voice: Voice label if known
• speaker_id: Unique speaker identifier

Example format:

{
  "text": "M ba Soaamb da mi n kėnd n tãooda ware...",
  "duration": 4.85,
  "voice": "female",
  "speaker_id": "2",
  "audio": { "path": "...", "array": [...], "sampling_rate": 16000 },
  "denoised_audio": { "path": "...", "array": [...], "sampling_rate": 16000 }
}

Dataset Overview

• 📂 Split: train
• 🔢 Total examples: 5535
• 🕒 Total duration: ~8.66 hours

---

Use Cases

This dataset is designed for:

• Training or evaluating ASR and TTS systems in Mooré
• Research on low-resource language modeling
• Pretraining for multilingual or African language models
• Preservation of oral cultural heritage in digital format

❌ Commercial use is strictly prohibited.

---

Access Policy (Important)

The dataset is under gated mode to ensure ethical use and protect cultural data integrity. To request access, please contact Wendyellé A. Alban NYANTUDRE ([email protected]).

Access will only be granted to individuals or institutions who:

• Clearly identify themselves
• Describe their intended academic or research use
• Agree to the non-commercial usage restriction

🔐 For safety and traceability reasons, anonymous or vague requests will not be approved.

---

Contributing or Collaborating

We welcome collaboration from anyone passionate about representing Mooré in language technologies! There are many ways to contribute:

• Suggest improvements or extensions to the dataset
• Share scripts, tools, or preprocessing tips in the GitHub repo (Moore Speech Corpora Toolkit ⭐)
• Improve documentation or metadata
• Help with validation, alignment, or text normalization

✅ Just make sure your contribution is meaningful, reproducible, and documented.
Together, let’s help Mooré shine in the world of AI! 💛
To contribute, feel free to open an issue or contact us directly.

---

License

The original audio/text content is sourced from mooreburkina.com and this processed dataset is licensed under CC BY-NC 4.0, allowing attribution-required, non-commercial use.

• Original audio/text: Copyright MooreBurkina.com – All rights reserved
• This processed dataset: Licensed under CC BY-NC 4.0 (Attribution required; strictly non-commercial use only)

For commercial or other usage scenarios not covered by this license, please contact us. By accessing this dataset, you agree to these terms.

---

Citation

@misc{nyantudre2025moorespeechcorpora,
  title        = {Moore Speech Corpora: A Cleaned, Denoised Audio-Text Dataset for Mooré TTS and ASR},
  author       = {Nyantudre, Wendyellé Abubakrh Alban and Bandaogo, Aristide},
  year         = {2025},
  howpublished = {Hugging Face Datasets (gated access)}
}

---

Contacts

• Curator: Wendyellé A. Alban NYANTUDRE

  • 📧 [email protected]
  • 💼 LinkedIn

• Aristide BANDAOGO

  • 📧 [email protected]