Ryota Komatsu

Interests

Speech-text large language models, reinforcement learning

Education

Apr 2025 – Mar 2028	PhD student, Institute of Science Tokyo (formerly Tokyo Institute of Technology) Specialization: Spoken language processing
Mar 2023	M.Eng., Institute of Science Tokyo (formerly Tokyo Institute of Technology) Specialization: Spoken language processing
Mar 2021	B.Eng., Institute of Science Tokyo (formerly Tokyo Institute of Technology) Specialization: Mathematical optimization

Work experience

Apr 2023 – Jan 2025

Researcher, Hitachi, Ltd.

selected publications

OJSP

Speaker-Disentangled Chunk-Wise Regression for Syllabic Tokenization

Ryota Komatsu, Kota Kawakita, Takuma Okamoto, and 1 more author

IEEE Open Journal of Signal Processing, 2026

arXiv Bib Code Website

@article{Komatsu_SylReg_2026,
  author = {Komatsu, Ryota and Kawakita, Kota and Okamoto, Takuma and Shinozaki, Takahiro},
  title = {Speaker-Disentangled Chunk-Wise Regression for Syllabic Tokenization},
  year = {2026},
  volume = {7},
  journal = {IEEE Open Journal of Signal Processing},
  pages = {800--808},
}

JSTSP

Automatic Spoken Language Acquisition Based on Observation and Dialogue

Ryota Komatsu, Shengzhou Gao, Wenxin Hou, and 9 more authors

IEEE Journal of Selected Topics in Signal Processing, 2022

Bib PDF Code

@article{9817627,
  author = {Komatsu, Ryota and Gao, Shengzhou and Hou, Wenxin and Zhang, Mingxin and Tanaka, Tomohiro and Toyoda, Keisuke and Kimura, Yusuke and Hino, Kent and Iwamoto, Yu and Mori, Kosuke and Okamoto, Takuma and Shinozaki, Takahiro},
  journal = {IEEE Journal of Selected Topics in Signal Processing},
  title = {Automatic Spoken Language Acquisition Based on Observation and Dialogue},
  year = {2022},
  volume = {16},
  number = {6},
  pages = {1480-1492},
}