---
license: cc-by-4.0
language:
- zh
- en
tags:
- audio
- speech
---

### Dataset Description

This dataset contains 93 Cantonese sentences, which include some code-mixing with English, that reflect everyday spoken language in Hong Kong. The sentences cover topics such as daily life, food, weather, leisure, personal feelings, and common social interactions. The former part of the dataset mainly focuses on simple daily dialogues and question–answer exchanges, while the latter part features proverbs, slang, idioms, and emotionally expressive speech that show greater variation in tone and context. Overall, the dataset represents relatively slow, deliberate speech, as the sentences are short, clearly segmented, and use full expressions.

### Issues Encountered & Solution

I encountered background noise from the traffic while recording. I know that over-removing background noise may accidentally remove some useful "noise" (e.g., fricatives). Therefore, I put myself in a closed room to minimize the damage from the traffic. And I use the reduce noise function with -10.0(dB) in Praat during the duration of the noise.I also encountered a situation where there are too many pauses in my stuttered recordings. I used ctrl + x to cut some of the highlighted and unwanted pauses to make the audio files more natural and as clean as possible.