[ 
| title: | RECNET - the speech recognition system |
| author: | J. Bittner, L.J.M. Rothkrantz & M. Snorek |
| published in: | 1996 |
| appeared as: |
Report of the Faculty of Technical Mathematics and Informatics nr. 96-160 Faculty of Technical Mathematics and Informatics Delft University of Technology |
| also as: |
Alparon report nr. 96-06 Section of Knowledge Based Systems Faculty of Technical Mathematics and Informatics Delft University of Technology |
Abstract
This paper presents some practical experiences with RECNET - continuous speech recognition system, as well as some theoretical background.
First chapter is a brief introduction to fundamentals of speech recognition and was added just for completeness.
Second chapter describes theoretically the RECNET system. It is based on articles of Tony Robinson - the author of RECNET.
Chapter 3 deals with the problem of adapting RECNET to training on various platforms, especially on nCUBE supercomputer. It contains description of parallel nCUBE implementations and presents some results of the training process.
The Appendix contains description of the extended version of RECNET v.1.3, called RECNET-E. Overview of file formats, executables and scripts, together with some examples of their usage is presented.
