SBCSAE / README.md
dklement's picture
Update README.md
d14a732 verified
metadata
dataset_info:
  features:
    - name: id
      dtype: string
    - name: duration
      dtype: float32
    - name: audio
      dtype: audio
    - name: transcript
      dtype: string
    - name: asr_aligned_transcript
      dtype: string
    - name: diar_aligned_transcript
      dtype: string
    - name: segments
      sequence:
        - name: start
          dtype: float32
        - name: end
          dtype: float32
        - name: speaker
          dtype: string
        - name: transcript
          dtype: string
    - name: asr_aligned_segments
      sequence:
        - name: start
          dtype: float32
        - name: end
          dtype: float32
        - name: speaker
          dtype: string
        - name: transcript
          dtype: string
    - name: diar_aligned_segments
      sequence:
        - name: start
          dtype: float32
        - name: end
          dtype: float32
        - name: speaker
          dtype: string
        - name: transcript
          dtype: string
  splits:
    - name: test
      num_bytes: 7186749331
      num_examples: 60
  download_size: 7047921491
  dataset_size: 7186749331
configs:
  - config_name: default
    data_files:
      - split: test
        path: data/test-*
license: mit
task_categories:
  - automatic-speech-recognition
language:
  - en

A detailed dataset description (including description, audio samples, and statistics) is provided here: https://domklement.github.io/sbcsae/

If you use the dataset, please, do not forget to cite our work:

@inproceedings{maciejewski24_interspeech,
  title     = {Evaluating the Santa Barbara Corpus: Challenges of the Breadth of Conversational Spoken Language},
  author    = {Matthew Maciejewski and Dominik Klement and Ruizhe Huang and Matthew Wiesner and Sanjeev Khudanpur},
  year      = {2024},
  booktitle = {Interspeech 2024},
  pages     = {2155--2159},
  doi       = {10.21437/Interspeech.2024-2119},
  issn      = {2958-1796},
}

Description of the HuggingFace dataset is coming soon.