korebaju_corpus / README.md
YuDZEN's picture
[update] README.md
da922da
metadata
license: mit
language:
  - coe
tags:
  - endangered language
  - korebaju

This is a Korebaju language dataset under construction.

check https://github.com/YuDZEN/korebaju-ASR for the Kaldi and MFA project for the same project

Audio Data

Audio data is in the audio folder. The audio data is in the wav format.

Transcription Data

Transcription data is test.jsonl and train.jsonl in the racine folder. The transcription data is in the jsonl format.

Data Split

The data is split into train and test sets. The train set contains 90% of the data and the test set contains 10% of the data.

data augmentation

We used data augmentation in order to increase the diversity of our training data without actually collecting new data (we have a very small dataset). By applying noise and volume, we can generate some training samples, which are named by "_NOISE" or "_VOLUME" after speakers' names.