Dataset · edinburghcstr
AMI meeting corpus: 100 hours of audio
edinburghcstr/ami
100 hours of meeting recordings with synchronized microphones, video, and text for ASR and meeting analysis.
What's inside
The AMI Meeting Corpus features 100 hours of English meeting recordings. The corpus synchronizes close and distant microphones, individual and general video cameras, projector and whiteboard outputs, and pen recordings. This is a classic resource for ASR and multimodal meeting analysis.
What you can do with it
- Training and evaluating ASR for meetings
- Speaker diarization
- Meeting transcription
- Tasks
- Automatic speech recognition
- Languages
- English
- Size
- 100K–1M rows
- Downloads this month
- 48.5K
- Likes
- 100
Models trained on it
- arbazsiddiqui/voxlockAudio → audio 2.6K
- BUT-FIT/SE-DiCoWAutomatic speech recognition 1.6K
- BUT-FIT/DiCoW_v3_3Automatic speech recognition 1.1K
- futo-org/asr4all-lAutomatic speech recognition 781
- futo-org/asr4all-sAutomatic speech recognition 587
- futo-org/asr4all-mAutomatic speech recognition 458
- mradermacher/tiny-audio-GGUF 434
- oruk/oruk-loop-153MAutomatic speech recognition 374
How to load it
from datasets import load_dataset
ds = load_dataset("edinburghcstr/ami")