Dataset · anyantudre
Cleaned Moore speech corpus for ASR and TTS
anyantudre/MooreSpeechCorporaV2
Between 100k and 1M audio-text pairs in Moore with verified alignment for ASR and TTS.
What's inside
The second version of the MooreSpeechCorpora dataset contains between 100k and 1M audio recordings in the low-resource Moore language (with French mentioned in metadata) paired with text. All clips were re-aligned, edge music was trimmed, speakers were re-labeled, and transcriptions were verified by two ASR models. The structure of the first six columns is fully compatible with the first version of the dataset.
What you can do with it
- Training and fine-tuning automatic speech recognition (ASR) models for Moore
- Training text-to-speech (TTS) systems using speaker annotations
- Tasks
- Automatic speech recognition, Text → speech
- Languages
- Mossi, French
- Size
- 100K–1M rows
- Downloads this month
- 35.6K
- Likes
- 0
- Access
- on request on Hugging Face
How to load it
from datasets import load_dataset
ds = load_dataset("anyantudre/MooreSpeechCorporaV2")