EVERA: an evolutionary programming environment for adaptive speech processing. n this paper, we describe EVERA (Environnement de Vie artificielle pour l’Etude de la Robustesse des Apprentis), a system for the modeling and simulation of adaptive speech processing systems. The possibility to combine genetic algorithms and artificial neural networks (ANNs) in the EVERA system allows the study of “evolutive” speech recognition systems. EVERA is written in an object-oriented framework and provides a high-level graphic user interface to design, observe and to monitor different simulations. We present some applications of the system to the problem of robust speech perception and recognition in hostile synthetic acoustic environments built from some physical model of real rooms (4 walls, 1 floor and 1 ceiling) which include different real world noises.
References in zbMATH (referenced in 1 article , 1 standard article )
Showing result 1 of 1.
- Kabré, Harouna; Spalanzani, Anne: EVERA: an evolutionary programming environment for adaptive speech processing (2000)