Dataset · basis-ai
Multilingual multi-speaker conversations (48 kHz)
basis-ai/basis-conversations-1500
Multilingual dataset of live dialogues of up to 4 people with separate audio channels for ASR.
What's inside
The dataset contains multi-party dialogues in various languages (including English, German, French, Spanish, Arabic, and Hebrew) involving two to four speakers simultaneously. Audio is recorded at 48 kHz with a separate channel for each speaker; the median conversation duration is 33 minutes, featuring a total of 2,645 unique speakers.
What you can do with it
- Train and evaluate ASR systems for overlapping dialogues (full-duplex)
- Develop speaker diarization and multi-speaker speech separation models
- Train voice AI assistants for natural real-time communication
- Tasks
- Automatic speech recognition, Audio classification
- Languages
- Arabic, German, English, Spanish, French, Hebrew, Hindi, Italian and 4 more
- Size
- 10K–100K rows
- Downloads this month
- 14.4K
- Likes
- 63
- Access
- on request on Hugging Face
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("basis-ai/basis-conversations-1500")