High-Accuracy Large-Vocabulary Speech Recognition Using Mixture Tying and Consistency Modeling

Abstract

Improved acoustic modeling can significantly decrease the error rate in large-vocabulary speech recognition. Our approach to the problem is twofold. We first propose a scheme that optimizes the degree of mixture tying for a given amount of training data and computational resources. Experimental results on the Wall Street Journal (WSJ) Corpus show that this new form of output distribution achieves a 25% reduction in error rate over typical tied- mixture systems. We then show that an additional improvement can be achieved by modeling local time correlation with linear discriminant features.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1994
Accession Number
ADA460273

Entities

People

  • Hy Murveit
  • Vassilios Digalakis

Organizations

  • SRI International

Tags

Communities of Interest

  • Advanced Electronics
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Automated Speech Recognition
  • Clustering
  • Coefficients
  • Consistency
  • Errors
  • Gaussian Distributions
  • Hidden Markov Models
  • Language
  • Markov Models
  • Models
  • Natural Languages
  • Probability
  • Recognition
  • Vocabulary
  • Word Recognition

Fields of Study

  • Computer science

Readers

  • Computational Linguistics
  • Computational Modeling and Simulation
  • Computer Vision.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Neural Networks