Conference proceeding
Discriminative segmental cascades for feature-rich phone recognition
2015 IEEE Workshop on Automatic Speech Recognition and Understanding : ASRU 2015 : proceedings : December 13-17, 2015, pp.561-568
12/01/2015
DOI: 10.1109/ASRU.2015.7404845
Abstract
Discriminative segmental models, such as segmental conditional random fields (SCRFs) and segmental structured support vector machines (SSVMs), have had success in speech recognition via both lattice rescoring and first-pass decoding. However, such models suffer from slow decoding, hampering the use of computationally expensive features, such as segment neural networks or other high-order features. A typical solution is to use approximate decoding, either by beam pruning in a single pass or by beam pruning to generate a lattice followed by a second pass. In this work, we study discriminative segmental models trained with a hinge loss (i.e., segmental structured SVMs). We show that beam search is not suitable for learning rescoring models in this approach, though it gives good approximate decoding performance when the model is already well-trained. Instead, we consider an approach inspired by structured prediction cascades, which use max-marginal pruning to generate lattices. We obtain a high-accuracy phonetic recognition system with several expensive feature types: a segment neural network, a second-order language model, and second-order phone boundary features.
Details
- Title: Subtitle
- Discriminative segmental cascades for feature-rich phone recognition
- Creators
- Hao Tang - Toyota Technological Institute at ChicagoWeiran Wang - Toyota Technological Institute at ChicagoKevin Gimpel - Toyota Technological Institute at ChicagoKaren Livescu - Toyota Technological Institute at Chicago
- Resource Type
- Conference proceeding
- Publication Details
- 2015 IEEE Workshop on Automatic Speech Recognition and Understanding : ASRU 2015 : proceedings : December 13-17, 2015, pp.561-568
- DOI
- 10.1109/ASRU.2015.7404845
- eISBN
- 1479972916; 9781479972913
- Language
- English
- Date published
- 12/01/2015
- Academic Unit
- Computer Science
- Record Identifier
- 9984696716902771
Metrics
1 Record Views