ouktlab/t5_sct-jis-v1_corpus10-bccwj-wiki40b_std

T5 SCT model

ouktlab/t5_sct-jis-v1_corpus10-bccwj-wiki40b_std

This is a Japanese syllable-to-character translation (SCT) model for character (Kanji, Katakana and Hiragana) recognition.

• This model is based on T5 architecture.
• This Japanese character tokenizer (v1) based on JIS X 0213 is assumed.
• The details and examples are shown at our github repository.

Citations

@inproceedings {rtakeda2025:apsipa,
  author={Ryu Takeda and Kazunori Komatani},
  title={Reducing Orthographic Dependency on Paired Data by Probabilistic Integration via Syllabogram for Japanese Dialogue Speech Recognition},
  year={2025},
  booktitle={Proceedings of Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) (to appear)},
}