Welcome to DataLabX, a research-focused team dedicated to building open, high-quality datasets for speech and language applications.
You can find our published datasets on Hugging Face 🤗.
- Language: English (speech + transcript)
- Size: ~10k samples, ~20 hours
- Use Case: Automatic Speech Recognition (ASR)
- Language: Japanese (audio) → Chinese (translation)
- Size: ~10k samples, ~30 hours
- Use Case: Speech-to-Text Translation
All datasets are released under CC BY 4.0, unless otherwise specified.
If you're interested in collaborating or using our datasets in your work, feel free to reach out on Hugging Face or submit issues and suggestions in this repo.
Maintained by ItBanque