Metraipolis › archive › dc_ed0ee374b108

Scaling Forced Alignment to End-User Devices — arXiv:2609.21145

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.21145
License (as archived)
CC0-1.0
Archived
2026-10-10T01:27:11.965Z
SHA-256 of the archived text
5f07301c83ebf29897bca9338ff2670a490c4aba1c1b7098f1f394a5a4c6984d
Stable URL of this page
https://metraipolis.com/documento/dc_ed0ee374b108

How to cite

Scaling Forced Alignment to End-User Devices — arXiv:2609.21145 — arXiv (en), https://arxiv.org/abs/2609.21145 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ed0ee374b108 (Metraipolis archive dc_ed0ee374b108, sha256 5f07301c83ebf29897bca9338ff2670a490c4aba1c1b7098f1f394a5a4c6984d)
BibTeX
@misc{dced0ee374b108,
  title = {Scaling Forced Alignment to End-User Devices — arXiv:2609.21145},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.21145},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ed0ee374b108; sha256 5f07301c83ebf29897bca9338ff2670a490c4aba1c1b7098f1f394a5a4c6984d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21145. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

The Viterbi algorithm has been previously used to perform forced alignment of audio to text to mine training data from online resources. However, many existing implementations have quadratic time and space complexity, scaling poorly to long input sequences. We propose two optimizations to address this issue. First, we apply the Hirschberg algorithm to perform the alignment in place using linear memory. Second, we model the alignment between speech and text as a constrained random walk, allowing us to prune the search space with arbitrary confidence while accounting for transcription errors. The Hirschberg optimization reduces memory usage from 140 GB to 5 MB for three-hour inputs while producing identical alignments in one-third the time of torchaudio when both run on a CPU. We achieve an additional 2x speedup with pruning on inputs longer than 20 minutes while preserving alignment accuracy in more than 98% of tested cases.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Data Structures and Algorithms (cs.DS).
Pagina del lavoro: https://arxiv.org/abs/2609.21145 — 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 2026-10-10T01:00:30.266Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).