Metraipolis › archive › dc_c43b696a63bb

Towards Efficient Simultaneous Inverse Text Normalization with Pretrained Text-to-Text Language Model and Read-Tag-Write Policy — arXiv:2610.19492

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.19492
License (as archived)
CC0-1.0
Archived
2026-10-15T20:23:55.142Z
SHA-256 of the archived text
32e88df8e583b21323ee2983f7bf941e4025cfc459b3354efb9bd8292cdc639f
Stable URL of this page
https://metraipolis.com/documento/dc_c43b696a63bb

How to cite

Towards Efficient Simultaneous Inverse Text Normalization with Pretrained Text-to-Text Language Model and Read-Tag-Write Policy — arXiv:2610.19492 — arXiv (en), https://arxiv.org/abs/2610.19492 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c43b696a63bb (Metraipolis archive dc_c43b696a63bb, sha256 32e88df8e583b21323ee2983f7bf941e4025cfc459b3354efb9bd8292cdc639f)
BibTeX
@misc{dcc43b696a63bb,
  title = {Towards Efficient Simultaneous Inverse Text Normalization with Pretrained Text-to-Text Language Model and Read-Tag-Write Policy — arXiv:2610.19492},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.19492},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c43b696a63bb; sha256 32e88df8e583b21323ee2983f7bf941e4025cfc459b3354efb9bd8292cdc639f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19492. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Streaming Inverse Text Normalization (ITN) is vital for converting spoken-form outputs from streaming Automatic Speech Recognition into formatted written text. Existing streaming ITN methods rely on hybrid systems combining neural tagging with expert-crafted finite-state transducer rules, limiting scalability across domains and languages. While end-to-end models offer superior scalability by learning directly from data, standard encoder-decoder architectures are inherently non-streaming due to global attention. In this paper, we propose an efficient streaming end-to-end ITN system, adapting a pretrained text-to-text model to leverage its robust linguistic knowledge. To enable streaming, we introduce architectural adaptations, a specialized training strategy, a Read-Tag-Write decoding policy, and inference optimizations. Experiments on Vietnamese datasets show accuracy comparable to non-streaming baselines, outperforming hybrid methods while satisfying real-time latency requirements.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.19492 — 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 2026-10-15T19:36:18.196Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).