Robust Models and Features for Speech Recognition.

Abstract

The goal of this project were: (1) to develop algorithms for achieving significant improvements in accuracy over current state-of-the-art speech recognition systems, (2) to implement these algorithms and assemble them into a prototype for a hands-on demonstration and (3) to test these algorithms for the transcription of "real-world" speech like radio/TV broadcast shows.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 13, 1998
Accession Number
ADA344834

Entities

People

  • David Nahamoo

Organizations

  • IBM Thomas J. Watson Research Center

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Accuracy
  • Adaptive Training
  • Algorithms
  • Automated Speech Recognition
  • Background Noise
  • Computer Vision
  • Decoding
  • Hidden Markov Models
  • Language
  • Markov Models
  • Models
  • Probability
  • Probability Distributions
  • Recognition
  • Signal Processing
  • Training
  • Vocabulary

Fields of Study

  • Computer science

Readers

  • Instructional Design and Training Evaluation.
  • Speech Processing/Speech Recognition.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks