@misc{indiciae1278b120994e,
  title = {Speaker-Disentangled Chunk-Wise Regression for Syllabic Tokenization},
  author = {Ryota Komatsu and Kota Kawakita and Takuma Okamoto and Takahiro Shinozaki},
  year = {2026},
  url = {https://arxiv.org/abs/2607.04064},
  note = {Source identifier: 2607.04064}
}