Datasets:

Modalities:
Audio
Languages:
Chinese
License:
11607279_LeeLeeYan / README.md
lillianllyan's picture
Update README.md
d283325 verified
metadata
license: cc-by-4.0
language:
  - zh
tags:
  - audio
  - speech

Dataset Description

This data set contains 57 everyday Cantonese phrases from Hong Kong, mainly covering topics such as work, school, food, and sports. The data set includes simple daily dialogues and Q&A sessions, as well as Cantonese phonetic transcriptions. This allows people unfamiliar with Cantonese pronunciation to learn the fun aspects of Cantonese through these transcriptions. Overall, this data set reflects the brevity, clear phrasing, and complete expression of Cantonese sentences, as well as the interesting aspects of spoken Cantonese.

Issues Encountered & Solution

During the recording, there are some background noises which makes my recording sounds not good enough. However, I use the "Reduce noise" function of praat to reduce the extra noise, so that my audio could sounds better.