VSASV-Dataset / README.md
hustep-admin's picture
Update README.md
549b4cf verified
metadata
license: cc-by-4.0
dataset_info:
  features:
    - name: file
      dtype: string
    - name: audio
      struct:
        - name: array
          sequence: float64
        - name: sampling_rate
          dtype: int64
    - name: label
      dtype: string
    - name: utt_type
      dtype: string
  splits:
    - name: train
      num_bytes: 215847527002
      num_examples: 220963
  download_size: 69218972580
  dataset_size: 215847527002
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
language:
  - vi
tags:
  - speaker-verification
  - speaker-recognition
  - spoofing-aware-speaker-verification
  - vietnamese
pretty_name: VSASV Dataset
size_categories:
  - 100K<n<1M

This paper introduces the VSASV dataset for Spoofing-Aware Speaker Verification (SASV) in Vietnamese language. A part of this dataset was used to host the Spoofing-aware Speaker Verification Challenge in VLSP 2023.

The paper has been accepted at INTERSPEECH 2024.