End-to-end training approaches for discriminative segmental models

Hao Tang; Weiran Wang; Kevin Gimpel; Karen Livescu

doi:10.1109/SLT.2016.7846309

Back

Conference proceeding

End-to-end training approaches for discriminative segmental models

Hao Tang, Weiran Wang, Kevin Gimpel and Karen Livescu

2016 IEEE Spoken Language Technology Workshop (SLT), pp.496-502

12/2016

DOI: 10.1109/SLT.2016.7846309

View Online

Abstract

Recent work on discriminative segmental models has shown that they can achieve competitive speech recognition performance, using features based on deep neural frame classifiers. However, segmental models can be more challenging to train than standard frame-based approaches. While some segmental models have been successfully trained end to end, there is a lack of understanding of their training under different settings and with different losses.

Computational modeling

Discriminative segmental models

end-to-end training

Fasteners

Hidden Markov models

Mel frequency cepstral coefficient

Neural networks

Predictive models

Training

Details

Title: Subtitle: End-to-end training approaches for discriminative segmental models
Creators: Hao Tang - Toyota Technological Institute at Chicago
Weiran Wang - Toyota Technological Institute at Chicago
Kevin Gimpel - Toyota Technological Institute at Chicago
Karen Livescu - Toyota Technological Institute at Chicago
Resource Type: Conference proceeding
Publication Details: 2016 IEEE Spoken Language Technology Workshop (SLT), pp.496-502
Publisher: IEEE
DOI: 10.1109/SLT.2016.7846309
Language: English
Date published: 12/2016
Academic Unit: Computer Science
Record Identifier: 9984696720202771

Metrics

1 Record Views