A distributed system architecture for speech recognition

  • Fil Alleva ,
  • Roberto Bisiani ,
  • S. Forin ,
  • Richard A. Lerner

IEEE International Conference on ICASSP '86 Acoustics, Speech, and Signal Processing |

Publication

This paper is an overview of the software environment (called AGORA) that is being developed at Carnegie-Mellon University to support the devlopment of connected, speaker independent, speech recognition systems. This includes the design, implemenentation, testing and fast (parallel) execution of speech recognition programs composed of a large number of independent algorithms. Systems described within AGORA can be executed on a number of different multiprocessor configurations.