A Comparison Between Deep Neural Nets and ...
Type de document :
Communication dans un congrès avec actes
Titre :
A Comparison Between Deep Neural Nets and Kernel Acoustic Models for Speech Recognition
Auteur(s) :
Lu, Zhiyun [Auteur]
University of California [Los Angeles] [UCLA]
Guo, Dong [Auteur]
University of Southern California [USC]
Garakani, Alireza Bagheri [Auteur]
University of Southern California [USC]
Liu, Kuan [Auteur]
University of Southern California [USC]
May, Avner [Auteur]
Bellet, Aurelien [Auteur]
Machine Learning in Information Networks [MAGNET]
Fan, Linxi [Auteur]
University of Southern California [USC]
Collins, Michael [Auteur]
Kingsbury, Brian [Auteur]
IBM Thomas J. Watson Research Center
Picheny, Michael [Auteur]
IBM Thomas J. Watson Research Center
Sha, Fei [Auteur]
University of California [Los Angeles] [UCLA]
University of California [Los Angeles] [UCLA]
Guo, Dong [Auteur]
University of Southern California [USC]
Garakani, Alireza Bagheri [Auteur]
University of Southern California [USC]
Liu, Kuan [Auteur]
University of Southern California [USC]
May, Avner [Auteur]
Bellet, Aurelien [Auteur]
Machine Learning in Information Networks [MAGNET]
Fan, Linxi [Auteur]
University of Southern California [USC]
Collins, Michael [Auteur]
Kingsbury, Brian [Auteur]
IBM Thomas J. Watson Research Center
Picheny, Michael [Auteur]
IBM Thomas J. Watson Research Center
Sha, Fei [Auteur]
University of California [Los Angeles] [UCLA]
Titre de la manifestation scientifique :
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2016)
Ville :
Shanghai
Pays :
Chine
Date de début de la manifestation scientifique :
2016-03-20
Mot(s)-clé(s) en anglais :
deep neural networks
kernel methods
acoustic models
automatic speech recognition
kernel methods
acoustic models
automatic speech recognition
Discipline(s) HAL :
Informatique [cs]/Apprentissage [cs.LG]
Statistiques [stat]/Machine Learning [stat.ML]
Statistiques [stat]/Machine Learning [stat.ML]
Résumé en anglais : [en]
We study large-scale kernel methods for acoustic modeling and compare to DNNs on performance metrics related to both acoustic modeling and recognition. Measuring perplexity and frame-level classification accuracy, kernel-based ...
Lire la suite >We study large-scale kernel methods for acoustic modeling and compare to DNNs on performance metrics related to both acoustic modeling and recognition. Measuring perplexity and frame-level classification accuracy, kernel-based acoustic models are as effective as their DNN counterparts. However, on token-error-rates DNN models can be significantly better. We have discovered that this might be attributed to DNN's unique strength in reducing both the perplexity and the entropy of the predicted posterior probabilities. Motivated by our findings, we propose a new technique, entropy regularized perplexity, for model selection. This technique can noticeably improve the recognition performance of both types of models, and reduces the gap between them. While effective on Broadcast News, this technique could be also applicable to other tasks.Lire moins >
Lire la suite >We study large-scale kernel methods for acoustic modeling and compare to DNNs on performance metrics related to both acoustic modeling and recognition. Measuring perplexity and frame-level classification accuracy, kernel-based acoustic models are as effective as their DNN counterparts. However, on token-error-rates DNN models can be significantly better. We have discovered that this might be attributed to DNN's unique strength in reducing both the perplexity and the entropy of the predicted posterior probabilities. Motivated by our findings, we propose a new technique, entropy regularized perplexity, for model selection. This technique can noticeably improve the recognition performance of both types of models, and reduces the gap between them. While effective on Broadcast News, this technique could be also applicable to other tasks.Lire moins >
Langue :
Anglais
Comité de lecture :
Oui
Audience :
Internationale
Vulgarisation :
Non
Collections :
Source :
Fichiers
- https://hal.inria.fr/hal-01329772/document
- Accès libre
- Accéder au document
- http://arxiv.org/pdf/1603.05800
- Accès libre
- Accéder au document
- https://hal.inria.fr/hal-01329772/document
- Accès libre
- Accéder au document
- https://hal.inria.fr/hal-01329772/document
- Accès libre
- Accéder au document
- document
- Accès libre
- Accéder au document
- icassp16.pdf
- Accès libre
- Accéder au document
- 1603.05800
- Accès libre
- Accéder au document