metadata
license: cc-by-nc-4.0
configs:
- config_name: default
data_files:
- split: eval
path: data/eval-*
dataset_info:
features:
- name: id
dtype: string
- name: mix
dtype:
audio:
decode: false
- name: transcript
dtype: string
- name: language
dtype: string
- name: speech_source
dtype: string
splits:
- name: eval
num_examples: 10
Voice Focus Examples
Collection of examples for accurate foreground speaker transcription.
Details
- Curated by: Joschka Wohlgemuth
- Funded by: ai-coustics GmbH