Development of Real-Time Speech Recognition.

Abstract

This phase developed parallel processing techniques appropriate for real time speech recognition and demonstrated the feasibility of achieving real time recognition on the BBN Butterfly Parallel Processor. The speech recognition algorithm that is the focus of this effort is being developed for the Byblos speech recognition system. The multiprocessor used in the research, the Butterfly Parallel Processor, is a shared memory, MIMD machine. The algorithm was implemented using the Uniform System software methodology, a system that simplifies parallel programming without sacrificing efficiency. The algorithm is described, highlighting those portions critical to an efficient parallel implementation. Some of the problems encountered in trying to improve the efficiency are presented as well as the solutions to those problems. The algorithm is shown to achieve 79% processor utilization on a 97 node Butterfly Parallel Processor. This is equivalent to speedup by a factor of 77 over a single processor benchmark.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1987
Accession Number
ADA176258

Entities

People

  • Lynn Cosell
  • Michael A. Krasner
  • Owen Kimball
  • Richard Schwartz

Organizations

  • BBN Technologies

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Algorithms
  • Automated Speech Recognition
  • Computations
  • Computer Programming
  • Contracts
  • Efficiency
  • Lepidoptera
  • Multiprocessors
  • Naval Warfare
  • Parallel Computing
  • Parallel Processing
  • Parallel Processors
  • Recognition
  • Signal Processing
  • System Software
  • Vocabulary
  • Warfare

Fields of Study

  • Computer science
  • Engineering

Readers

  • Parallel and Distributed Computing.
  • Speech Processing/Speech Recognition.

Technology Areas

  • AI & ML