Hybrid language models for speech transcription

Luiza Orosanu 1 Denis Jouvet 1
1 PAROLE - Analysis, perception and recognition of speech
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
Abstract : This paper analyzes the use of hybrid language models for automatic speech transcription. The goal is to later use such an approach as a support for helping communication with deaf people, and to run it on an embedded decoder on a portable device, which introduces constraints on the model size. The main linguistic units considered for this task are the words and the syllables. Various lexicon sizes are studied by setting thresholds on the word occurrence frequencies in the training data, the less frequent words being therefore syllabified. A recognizer using this kind of language model can output between 62% and 96% of words (with respect to the thresholds on the word occurrence frequencies; the other recognized lexical units are syllables). By setting different thresholds on the confidence measures associated to the recognized words, the most reliable word hypotheses can be identified, and they have correct recognition rates between 70% and 92%.
Document type :
Conference papers
Liste complète des métadonnées

Cited literature [25 references]  Display  Hide  Download

https://hal.inria.fr/hal-01090478
Contributor : Denis Jouvet <>
Submitted on : Wednesday, December 3, 2014 - 3:49:00 PM
Last modification on : Tuesday, December 18, 2018 - 4:38:02 PM
Document(s) archivé(s) le : Monday, March 9, 2015 - 5:50:42 AM

File

articleIS2014-LM-version du 20...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01090478, version 1

Collections

Citation

Luiza Orosanu, Denis Jouvet. Hybrid language models for speech transcription. INTERSPEECH 2014, 15th Annual Conference of the International Speech Communication Association, Sep 2014, Singapour, Singapore. ⟨hal-01090478⟩

Share

Metrics

Record views

280

Files downloads

215