Neural Network Architectures for Pattern Recognition

作者: Françoise Fogelman Soulié

DOI: 10.1007/978-3-642-79119-2_12

关键词: Pattern recognition (psychology)Image (mathematics)Artificial neural networkSpeech processingOptical character recognitionModular designTime delay neural networkComputer scienceNeocognitronArtificial intelligence

摘要: It is argued that maximal benefit gained from complex Neural Network architectures. A formalism to train Multi Modular Architectures given. Examples of such MMAs for various applications in image and speech processing, time series prediction are shown, which illustrate the benefits using MMAs.

参考文章(36)
Hazem Bouattour, Francoise Fogelman Soulié, Emmanuel Viennet, Solving the Human Face Recognition Task using Neural Nets international conference on artificial neural networks. pp. 1595- 1598 ,(1992) , 10.1016/B978-0-444-89488-5.50164-0
Jérôme Loncelle, Nicolas Derycke, Françoise Fogelman Soulié, Optical Character Recognition and Cooperating Neural Networks techniques international conference on artificial neural networks. pp. 1591- 1594 ,(1992) , 10.1016/B978-0-444-89488-5.50163-9
Emmanuel Viennet, Francoise Fogelman Soulié, Scene segmentation using Multiresolution Analysis and MLP international conference on artificial neural networks. pp. 1599- 1602 ,(1992) , 10.1016/B978-0-444-89488-5.50165-2
M. de BOLLIVIER, P. GALLINARI, S. THIRIA, NEURAL NETS AND TASK DECOMPOSITION international conference on artificial neural networks. pp. 1655- 1658 ,(1991) , 10.1016/B978-0-444-89178-5.50158-5
Vladimir Naumovich Vapnik, Estimation of Dependences Based on Empirical Data ,(2010)