# English_Accent_DataSet

English Accent DataSet is a 79-hour speech dataset containing 23 different English accents.
The raw audio data files are from [VCTK](https://datashare.ed.ac.uk/handle/10283/2651), [EDACC](https://groups.inf.ed.ac.uk/edacc/) and [Voxpopuli](https://aclanthology.org/2021.acl-long.80/).

## Features

- **audio_id**: Unique identifier for each audio file.
- **audio**: The audio data.
- **raw_text**: The raw transcription.
- **gender**: Gender of the speaker.
- **speaker_id**: Identifier for the speaker.
- **accent**: Accent of the speaker.
- **duration**: Duration of the audio file.
- **split**: Split for training, validataion and test.

## Accents

- Dutch
- German
- Polish
- French
- Hungarian
- Finnish
- Romanian
- Slovak
- Spanish
- Italian
- Estonian
- Lithuanian
- Croatian
- Slovene
- English
- Scottish
- Irish
- NorthernIrish
- Indian
- Vietnamese
- Canadian
- American