On-the-Fly Text Retrieval for end-to-end ASR Adaptation

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F23%3APU149423" target="_blank" >RIV/00216305:26230/23:PU149423 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10095857" target="_blank" >https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10095857</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICASSP49357.2023.10095857" target="_blank" >10.1109/ICASSP49357.2023.10095857</a>

Jazyk výsledku

angličtina

Název v původním jazyce

On-the-Fly Text Retrieval for end-to-end ASR Adaptation

Popis výsledku v původním jazyce

End-to-end speech recognition models are improved by incorporat- ing external text sources, typically by fusion with an external lan- guage model. Such language models have to be retrained whenever the corpus of interest changes. Furthermore, since they store the entire corpus in their parameters, rare words can be challenging to recall. In this work, we propose augmenting a transducer-based ASR model with a retrieval language model, which directly retrieves from an external text corpus plausible completions for a partial ASR hy- pothesis. These completions are then integrated into subsequent pre- dictions by an adapter, which is trained once, so that the corpus of interest can be switched without incurring the computational over- head of retraining. Our experiments show that the proposed model significantly improves the performance of a transducer baseline on a pair of question-answering datasets. Further, it outperforms shallow fusion on recognition of named entities by about 7% relative; when the two are combined, the relative improvement increases to 13%

Název v anglickém jazyce

On-the-Fly Text Retrieval for end-to-end ASR Adaptation

Popis výsledku anglicky

End-to-end speech recognition models are improved by incorporat- ing external text sources, typically by fusion with an external lan- guage model. Such language models have to be retrained whenever the corpus of interest changes. Furthermore, since they store the entire corpus in their parameters, rare words can be challenging to recall. In this work, we propose augmenting a transducer-based ASR model with a retrieval language model, which directly retrieves from an external text corpus plausible completions for a partial ASR hy- pothesis. These completions are then integrated into subsequent pre- dictions by an adapter, which is trained once, so that the corpus of interest can be switched without incurring the computational over- head of retraining. Our experiments show that the proposed model significantly improves the performance of a transducer baseline on a pair of question-answering datasets. Further, it outperforms shallow fusion on recognition of named entities by about 7% relative; when the two are combined, the relative improvement increases to 13%

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of ICASSP 2023

ISBN

978-1-7281-6327-7

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

1-5

Název nakladatele

IEEE Signal Processing Society

Místo vydání

Rhodes Island

Místo konání akce

Rhodes Island, Greece

Datum konání akce

4. 6. 2023

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—