data_mini / README.md
yanjia0's picture
Create README.md
71059ff
metadata
task_categories:
  - token-classification
language:
  - zh
tags:
  - tone
  - pinyin
  - sentence
  - audio
size_categories:
  - 100M<n<1B

Dataset Details

  • Welcome to the Single-Speaker Mandarin Audio Dataset! This dataset is a curated subset extracted from a larger collection, focusing on audio recordings of a single speaker. Each audio file is accompanied by valuable linguistic annotations, including Pinyin transcriptions, tone information, and onset and offset details.

Dataset Description

  • Speaker: The dataset exclusively features recordings of a single Mandarin speaker, providing consistency for various linguistic analyses and applications.
  • Pinyin Transcriptions: Each audio file comes with a corresponding Pinyin transcription, offering a phonetic representation of the spoken Mandarin.
  • Tone Information: Tone annotations are included to capture the tonal characteristics of the spoken language. This feature is essential for tone-related studies and applications.
  • Onset and Offset Details: Precise information about the onset and offset of each audio segment is provided. This allows for accurate segmentation and analysis of the spoken content.

Dataset Sources [optional]

  • Subset of the Original Kaggle Dataset

Uses

  • Use for model evaluation or demo