Metraipolis › archive › dc_3abe373ef079

CoSTA: Compressed Speech Token Adaptation for Efficient Spoken Dialogue Models — arXiv:2610.21143

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.21143
License (as archived)
CC0-1.0
Archived
2026-10-16T09:19:39.607Z
SHA-256 of the archived text
c9666397e36e4120a03b0688e30704d3b76b519f4bec6ddb49fa4ee2cceb12f9
Stable URL of this page
https://metraipolis.com/documento/dc_3abe373ef079

How to cite

CoSTA: Compressed Speech Token Adaptation for Efficient Spoken Dialogue Models — arXiv:2610.21143 — arXiv (en), https://arxiv.org/abs/2610.21143 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3abe373ef079 (Metraipolis archive dc_3abe373ef079, sha256 c9666397e36e4120a03b0688e30704d3b76b519f4bec6ddb49fa4ee2cceb12f9)
BibTeX
@misc{dc3abe373ef079,
  title = {CoSTA: Compressed Speech Token Adaptation for Efficient Spoken Dialogue Models — arXiv:2610.21143},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.21143},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3abe373ef079; sha256 c9666397e36e4120a03b0688e30704d3b76b519f4bec6ddb49fa4ee2cceb12f9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.21143. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Spoken Dialogue Models (SDMs) typically share a speech-token space for generation and synthesis, leading to long autoregressive sequences and costly inference. Lower-rate speech tokenizers shorten sequences but may introduce compression-quality trade-offs and costly replacement without breaking this coupling. To address this limitation, we propose CoSTA, a compatibility-preserving post-hoc framework that decouples generation and synthesis token spaces while retaining the original speech tokenizer and decoder. CoSTA extends the generation token space with compressed representations via mean-initialized vocabulary extension, pronunciation adaptation, and dialogue adaptation, followed by deterministic expansion back to the original speech-token space before synthesis. On bilingual speech-dialogue benchmarks, CoSTA achieves approximately $1.7\times$ expanded speech-token generation speed, lower full-response latency, and higher throughput while preserving response quality. Analyses show more sustained interaction under limited budgets and a practical trade-off among compression, decoding efficiency, and quality. Further experiments demonstrate cross-architecture transfer and complementarity to low-rate tokenization.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Sound (cs.SD).
Pagina del lavoro: https://arxiv.org/abs/2610.21143 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T08:44:17.914Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).