eduhk-compling/s1160724_GoQuayAdrieneFernandez

Dataset Description

<! -- Provide a longer summary of what this dataset is, -->

About three minutes of English spoken speech, comprising roughly 50 sentences, were captured by one speaker in a calm setting for this dataset. In order to comply with typical text-to-speech dataset standards, the primary recording was first recorded as an M4A file before being converted to 16-bit, 44.1 kHz WAV format. To facilitate simple automated access and use for TTS training, the audio is divided into brief sentence-level segments, each of which is coupled with its precise transcription in a metadata.csv file, which includes the RP IPA transcription from the speech text.

Issues Encountered & Solution

Using the Audacity app, I have faced a few problems in splitting the audio clips into small sections, as this is my first time using this application. That is why I used extra time to find the button to split the clips and used extra time to tune the volume of the audio. I solved this problem by asking an AI tool to provide me with the steps on how to adjust the volume and split audio clips.