System Combination by Driven Decoding

Benjamin Lecouteux 1 Georges Linarès 1 Yannick Estève 2 J. Mauclair 2
1 LIA - Laboratoire Informatique d'Avignon
2 LIUM - Laboratoire d'Informatique de l'Université du Mans
Abstract : The combination of Automatic Speech Recognition (ASR) systems generally relies on a posteriori merge of system outputs or on a cross-adaptation. In this paper, we propose an integrated approach where the search of a primary system is driven by the outputs of a secondary one. This method allows to drive the primary system search by using the one-best hypotheses and the word posteriors gathered from the secondary system. Experiments are carried out within the experimental framework of the ESTER evaluation campaign [1]. Results show that the driven decoding algorithm significantly outper-forms the two single ASR systems (-8% of relative WER,-1.7% absolute). Finally, we investigate the interactions between driven decoding and cross-adaptations. The best cross-adaptation strategy in combination with the driven decoding process brings to a final absolute gain of about 1.9% WER.
Keywords : confidence measures broadcast news transcription decoding algorithms system combination
Type de document :
Communication dans un congrès
IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07 , Apr 2007, Honolulu, United States. 2007, Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on. 〈10.1109/ICASSP.2007.366919〉
Liste complète des métadonnées

Littérature citée [10 références]  Voir  Masquer  Télécharger
https://hal.archives-ouvertes.fr/hal-01318073
Contributeur : Bibliothèque Universitaire Déposants Hal-Avignon <>
Soumis le : jeudi 9 novembre 2017 - 09:24:29
Dernière modification le : vendredi 26 janvier 2018 - 10:46:26
Document(s) archivé(s) le : samedi 10 février 2018 - 12:36:23

Fichier

System_combination_by_driven_d...
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

UNIV-LEMANS | LIUM | LIUM-LST

Citation

Benjamin Lecouteux, Georges Linarès, Yannick Estève, J. Mauclair. System Combination by Driven Decoding. IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07 , Apr 2007, Honolulu, United States. 2007, Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on. 〈10.1109/ICASSP.2007.366919〉. 〈hal-01318073〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

96

Téléchargements de fichiers

7

Contact

support.ccsd.cnrs.fr
hal.support@ccsd.cnrs.fr
Logo CCSD