{"ID":5938068,"CreatedAt":"2026-07-07T03:14:33.014478982Z","UpdatedAt":"2026-07-13T23:59:41.726494204Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.04064","arxiv_id":"2607.04064","title":"Speaker-Disentangled Chunk-Wise Regression for Syllabic Tokenization","abstract":"Unsupervised syllabic tokenization aims to learn discrete syllabic tokens that capture latent linguistic content-related structure from raw speech. Recent syllabic tokenization methods employ teacher-student distillation of the pretrained HuBERT to organize latent speech frame representations into syllabic segments. However, when trained with an utterance-level cross-entropy objective, the model predicts speaker identity rather than linguistic content, thereby compromising the purity of syllabic tokens. To address this problem, we propose a speaker-disentangled syllabic tokenizer that regresses speaker-perturbed student representations toward clean teacher targets within fixed-length chunks. Experimental results demonstrate that our proposed method achieves state-of-the-art performance in syllable boundary detection and syllabic segment clustering. Moreover, a speech language model trained on our syllabic tokens achieves a 7% relative improvement in syntactic and semantic understanding over the phone-level SpiRit-LM.","short_abstract":"Unsupervised syllabic tokenization aims to learn discrete syllabic tokens that capture latent linguistic content-related structure from raw speech. Recent syllabic tokenization methods employ teacher-student distillation of the pretrained HuBERT to organize latent speech frame representations into syllabic segments. Ho...","url_abs":"https://arxiv.org/abs/2607.04064","url_pdf":"https://arxiv.org/pdf/2607.04064v1","authors":"[\"Ryota Komatsu\",\"Kota Kawakita\",\"Takuma Okamoto\",\"Takahiro Shinozaki\"]","published":"2026-07-05T00:39:57Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.AI\",\"cs.SD\",\"eess.AS\"]","methods":"[\"Language Model\",\"Generative Adversarial Network\"]","has_code":false}