articleAug 25, 2013GREEN OA

Sequence-discriminative training of deep neural networks

Brno University of Technology · University of Edinburgh · +1 more institution

Indexed incrossref

Abstract

Sequence-discriminative training of deep neural networks (DNNs) is investigated on a 300 hour American English conversational telephone speech task. Different sequence-discriminative criteria ndash;- maximum mutual information (MMI), minimum phone error (MPE), state-level minimum Bayes risk (sMBR), and boosted MMI ndash;- are compared. Two different heuristics are investigated to improve the performance of the DNNs trained using sequence-based criteria ndash;- lattices are re-generated after the first iteration of training; and, for MMI and BMMI, the frames where the numerator and denominator hypotheses are disjoint are removed from the gradient computation. Starting from a competitive DNN baseline trained…

Citation impact

662
total citations
FWCI
130.95
Percentile
100%
References
27
Citations per year

Authors

4

Topics & keywords

Keywords
  • Discriminative model
  • Training (meteorology)
  • Artificial intelligence
  • Computer science
  • Sequence (biology)
  • Artificial neural network
  • Pattern recognition (psychology)
  • Speech recognition
UN Sustainable Development Goals
  • Reduced inequalities
No related works found for this paper.

Funding