Title :
The implement and application of DBF neural networks
Author :
Wenming, Cao ; Hao, Feng ; Shoujue, Wang
Author_Institution :
Inst. of Intelligent Inf. Syst., Zhengjiang Univ. of Technol., Hangzhou, China
Abstract :
On the basis of DBF (direction basis function) nets proposed by Wang Shoujue et al. (see Proc. JCIS98, vol.IV, p.199-202, 1998), the model and implementation of a DBF neural network are discussed. When applied in pattern recognition, the algorithm and hardware implementation are presented. Compared with the traditional RBF, DBF neural networks are shown to offer advantages in accuracy, high speed and easy hardware implementation.
Keywords :
feedforward neural nets; pattern recognition; DBF neural networks; RBF neural networks; control systems; direction basis function nets; feedforward neural networks; image processing; pattern recognition; radial basis function neural networks; Computer aided analysis; Computer networks; Control systems; Feedforward neural networks; Kernel; Neural network hardware; Neural networks; Neurons; Pattern classification; Pattern recognition;
Conference_Titel :
Signal Processing, 2002 6th International Conference on
Print_ISBN :
0-7803-7488-6
DOI :
10.1109/ICOSP.2002.1180985
