Mike Rychener
Abstract
The most populous project at CMU is the speech recognition project headed by Professor Raj Reddy and Lee Erman. The current achievement is a voice-chess system incorporating several knowledge sources working in a hypothesize-and-test mode and interacting smoothly as a set of cooperating independent processes. This Hearsay system is reasonably successful in recognizing connected speech in the limited chess context, even though the chess-move grammar includes about five million possible utterances. It relies on a chess semantic specialist, the Tech program, on a grammar specialist, and on basic acoustic routines. Presently, re-analysis and reorganization of that effort is being undertaken, by Lee Erman, Victor Lesser, and Richard Fennell, with views towards implementation on the CMU C.mmp (multi-mini-processor) system. Bruce Lowerre and Richard Smith are analyzing the performance of Hearsay for multiple speakers on several other tasks: a Doctor task, voice news retrieval, and a desk calculator.
Recommendations
Investigation Amazigh speech recognition using CMU tools
The aim of this paper is to describe the development of a speaker-independent continuous automatic Amazigh speech recognition system. The designed system is based on the Carnegie Mellon University Sphinx tools. In the training and testing phase an in ...
Comments