---
language:
- yue
dataset_info:
  features:
  - name: audio_file
    dtype:
      audio:
        sampling_rate: 16000
  - name: duration
    dtype: int64
  - name: label
    dtype: string
  - name: confidence
    dtype: float64
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 22487193006.137
    num_examples: 726579
  download_size: 72192896899
  dataset_size: 22487193006.137
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
# Cantonese Emotional Speech

Crawled from YouTube and RTHK, this dataset contains 1,000 hours of Cantonese speech, each labeled with one of the following emotions: angry, disgusted, fearful, happy, neutral, other, sad, or surprised. The dataset also includes the confidence of the emotion label. The audio files are denoised with [resemble-enhance](https://github.com/resemble-ai/resemble-enhance). The transcriptions are generated by [SenseVoiceSmall](https://huggingface.co/FunAudioLLM/SenseVoiceSmall), and deduplicated using MinHash.