Title :
Multi-modular neural network-hybrid architectures: a review
Author :
Soulie, Françoise Fogelman
Author_Institution :
Lab. de Recherche en Inf., Univ. de Paris-Sud, Orsay, France
Abstract :
A formalism to train multi-modular architectures (MMAs) is given. Examples of MMAs for various applications in image and speech processing are shown, which illustrate the benefits of using MMAs.
Keywords :
face recognition; image recognition; learning (artificial intelligence); neural net architecture; neural nets; optical character recognition; parallel architectures; speech recognition; face recognition; image recognition; modular learning system; multi-modular architectures; multi-modular neural network; optical character recognition; speaker identification; speech recognition; Computer architecture; Cost function; Insulation; Lagrangian functions; Neural networks; Optical character recognition software; Optical computing; Speech processing; Speech recognition; Transfer functions;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.714170
