akumanatt / ko2kana

ko2kana

Tools for writing Korean pronunciation in katakana.

Replace Korean with Katakana so that you can read Korean in the TTS Japanese model. Typically, when you create a multilingual TTS, you organize the text with an IPA cleaner. However, this method is awkward, inaccurate and if the pronunciation symbol is insufficient or absent, the voice is missing. Therefore, only phonemes present in the TTS data were made close to the pronunciation of other languages.

Installation

pip install ko2kana

Usage

from ko2kana import toKana
s = toKana("안녕하세요.")
print(s) 
'アンニョンーハセヨ'

References

If you use our software for research, please cite:

@misc{kim2023ko2kana,
  author = {Kim, Gyeongmin},
  title = {ko2kana},
  year = {2023},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/kdrkdrkdr/ko2kana}}
}```