Published April 29, 2024 | Version v1
Dataset Restricted

Audio files from WhatTCSay 3

Creators

  • 1. Institut National des Langues et Civilisations Orientales

Description

This dataset contains wav files converted from the mp3 recorded for the WhatTCSay application.

It consists in 80 minutes of speech, reading 9146 syllables.

It was processed to serve as training data in Text To Speech experiments, and splitted into 4388 files used for training,
and 19 kept for testing. Results were published at LREC 2024 (Magistry, Wang & Lim, 2024)

 

Files

Restricted

The record is publicly accessible, but files are restricted to users with access.

Additional details

Additional titles

Subtitle (Min Nan Chinese)
Training Data for Teochew Text to Speech