DocumentCode :
324568
Title :
Using sparse trace neural network to extract temporal-spatial invariance
Author :
Peng, Hanchuan ; Gan, Qiang ; Sha, Lifeng ; Wei, Yu
Author_Institution :
Dept. of Biomed. Eng., Southeast Univ., Nanjing, China
Volume :
2
fYear :
1998
fDate :
4-9 May 1998
Firstpage :
1293
Abstract :
The relationship between invariance extraction and temporal processing in the visual system is far from being fully explored. A temporal-spatial invariance extraction model is proposed from the point of view of neuro-coding. Temporal self-adaptation is introduced into a neural invariance extractor to code some underlying physical parameters of input patterns and to learn traces of given classes of patterns. A trace is defined as the weighted temporal average over the output of the invariance extractor which leads to a simple trace neural network (TNN). A sparse trace neural network (STNN) is then proposed to produce pattern representations which can be easily post-processed. The behaviour of the STNN can be explained as forming some interesting temporal-spatial order structures and is compared qualitatively with some other models. Simulation results show that STNN achieves better performance than several previously proposed models
Keywords :
feature extraction; feedforward neural nets; invariance; learning (artificial intelligence); multilayer perceptrons; neurophysiology; physiological models; visual perception; input patterns; invariance extraction; neuro-coding; pattern representations; sparse trace neural network; temporal processing; temporal self-adaptation; temporal-spatial invariance; temporal-spatial order structures; weighted temporal average; Biomedical engineering; Brain modeling; Feedforward neural networks; Gallium nitride; Neural networks; Neurons; Radio frequency; Retina; Student members; Visual system;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks Proceedings, 1998. IEEE World Congress on Computational Intelligence. The 1998 IEEE International Joint Conference on
Conference_Location :
Anchorage, AK
ISSN :
1098-7576
Print_ISBN :
0-7803-4859-1
Type :
conf
DOI :
10.1109/IJCNN.1998.685961
Filename :
685961
Link To Document :
بازگشت