Dataset Description:
This dataset contains 50 Cantonese audio clips recorded by native Cantonese male speakers aged 20, along with their corresponding text transcriptions. The audio content covers daily Cantonese conversations, which introduce Hong Kong’s unique culture through local tea restaurants, famous locations, and idioms.
The audio dataset collection is approximately 3 minutes long, and it was recorded in a quiet and indoor space with the microphone of a MacBook Air. By Using the Praat… See the full description on the dataset page:
https://huggingface.co/datasets/eduhk-compling/11601706_QIUZHILIANG.
