README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: sentence
      dtype: string
  splits:
    - name: train
      num_bytes: 1333630236.308
      num_examples: 6219
  download_size: 1335711723
  dataset_size: 1333630236.308
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
language:
  - yue
license: cc-by-nc-nd-4.0

ASR-SCCantCabSC: A Scripted Chinese Cantonese (Canton) Cabin Speech Corpus

This open-source dataset consists of 5 hours of transcribed Guangzhou Cantonese scripted speech in the vehicle focusing on digits, commands and queries, where 6,219 utterances contributed by ten speakers were contained. Source: https://magichub.com/datasets/guangzhou-cantonese-scripted-speech-corpus-in-the-vehicle/