Investigating techniques for low resource conversational speech recognition - Le Mans Université Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Investigating techniques for low resource conversational speech recognition

Résumé

In this paper we investigate various techniques in order to build effective speech to text (STT) and keyword search (KWS) systems for low resource conversational speech. Sub-word decoding and graphemic mappings were assessed in order to detect out-of-vocabulary keywords. To deal with the limited amount of transcribed data, semi-supervised training and data selection methods were investigated. Robust acoustic features produced via data augmentation were evaluated for acoustic modeling. For language modeling, automatically retrieved conversational-like Webdata was used, as well as neural network based models. We report STT improvements with all the techniques, but interestingly only some improve KWS performance. Results are reported for the Swahili language in the context of the 2015 OpenKWS Evaluation.
Fichier principal
Vignette du fichier
Laurent16.pdf (152.65 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01515254 , version 1 (27-04-2017)

Identifiants

Citer

Antoine Laurent, Thiago Fraga-Silva, Lori Lamel, Jean-Luc Gauvain. Investigating techniques for low resource conversational speech recognition. 41st IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2016), Mar 2016, Shangai, China. pp.5975-5979, ⟨10.1109/ICASSP.2016.7472824⟩. ⟨hal-01515254⟩
165 Consultations
191 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More