Metraipolis › archive › dc_3fa31e0eda82

BaLEEN: Biasing with Latent Encoded Entities for Context-Aware ASR — arXiv:2609.36913

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.36913
License (as archived)
CC0-1.0
Archived
2026-10-07T16:26:30.546Z
SHA-256 of the archived text
b4154b1fc959592affd00e6934011fe4ea4e67afa9abe94a02c2957c94ad4b40
Stable URL of this page
https://metraipolis.com/documento/dc_3fa31e0eda82

How to cite

BaLEEN: Biasing with Latent Encoded Entities for Context-Aware ASR — arXiv:2609.36913 — arXiv (en), https://arxiv.org/abs/2609.36913 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3fa31e0eda82 (Metraipolis archive dc_3fa31e0eda82, sha256 b4154b1fc959592affd00e6934011fe4ea4e67afa9abe94a02c2957c94ad4b40)
BibTeX
@misc{dc3fa31e0eda82,
  title = {BaLEEN: Biasing with Latent Encoded Entities for Context-Aware ASR — arXiv:2609.36913},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36913},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3fa31e0eda82; sha256 b4154b1fc959592affd00e6934011fe4ea4e67afa9abe94a02c2957c94ad4b40},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36913. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Transcribing domain-specific entities and rare proper nouns remains a major challenge in automatic speech recognition (ASR). In this paper, we propose BaLEEN (Biasing with Latent Encoded Entities), a lightweight, hypernetwork-based framework for dynamic contextual adaptation without fine-tuning the underlying ASR model. BaLEEN encodes variable-length contextual keywords using a pretrained language model, compresses them into a fixed sequence of latent vectors via a Perceiver bottleneck, and injects context-dependent bias vectors directly into the intermediate encoder representations of the ASR model. Because both the language model and the backbone ASR model remain entirely frozen during training, BaLEEN operates as a plug-and-play adapter that incurs zero computational overhead at inference time when context biases are precomputed. We evaluate our method on a CTC-based ASR model using a Wikipedia-derived corpus with annotated named entities and synthetic speech. Experimental results demonstrate that BaLEEN reduces keyword miss rate by 8.7% on the test set relative to the unbiased baseline while simultaneously improving overall word error rate by 21% and character error rate by 28%.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.36913 — 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 2026-10-07T15:54:53.904Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).