人工神经网络FPGA实现研究进展与发展趋势

被引:6
作者
林祥金
张志利
朱智
机构
[1] 西安洪庆高技术研究所
关键词
人工神经网络; FPGA; 可重构技术; 智能计算;
D O I
10.14107/j.cnki.kzgc.2007.s3.009
中图分类号
TP183 [人工神经网络与计算];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
综述了人工神经网络FPGA实现的研究进展和关键技术,分析了如何利用FHGA的可重构技术来实现人工神经网络,探讨了实现过程中的一些问题,并介绍了作为神经网络FPGA实现的基础—可重构技术。指出测试平台设计、软件工具、FPGA友好学习算法及拓扑结构自适应等方面的研究,是今后研究的热点。
引用
收藏
页码:1 / 3
页数:3
相关论文
共 15 条
[1]  
Backpropagation a nalysis of the limited precision on high-order function neural networks. Jiang M,Gielen G. Proceedings of International Symposium on Neural Networks,August . 2004
[2]  
GangLion-a fast field programma-ble gate array im- plementation of a connectionist class-ier. Cox C E,Blanz E. IEEE Journal of Oceanic Engineering . 1992
[3]  
Partial and dynamic reconfiguration of FPGAs:a top down design methodology for an automatic implementation. Berthelot F,Nouvel F,Houzed D. IEEE Computer Society Annual symposium on emerging VLSI Technologies and Architectures . 2006
[4]  
Fast prototyping of artificial neural network:GSN digital implementa on. Simoes do Valle E,,Uebel L F,Barone D A. Proceed- ings of the Fourth International Conference on Microelectronics for Neu- ral Networks and Fuzzy Systems . 1994
[5]  
Density enhancement of a neural network using FPGAs and run-time reconfiguration. Eldredge J G,,Hutchings B L. Pro- ceedings of IEEE Workshop on FPGAs for Custom Computing Ma- chines . 1994
[6]  
A constant array multiplier core generator with dynamic partial evaluation architec ure selection. Yang B,,Joshi N,Karri R. 13th International Symposium on Field Programmable Gate Ar- rays . 2005
[7]  
Performance analysis of fixed,recon- figurable,and custom architectures for the SCAN image and video en- cryption. Dolla A,Kachris C,Bourbakis N. Proceedings of the 11th Annual IEEE Symposium on Field-Programmable Custom Computing Machines . 2003
[8]  
Dynamically self reconfigurable machine learnin structure for FPGA implementation. Starzyk J A,Guo Y. Proc Int Conf on Engineering of Reconfigurable Systems and Algorithms(ERSA) . 2003
[9]  
Neural network training with con- strained integer weights. Plagianakos V P,Vrahatis M N. Proceedings of the 1999 Congress on Evolutionary Computation . 1999
[10]  
Converting integer to floating point number. He J,,Li V,Han Y Q. Proceedings of the 2003 International Conference on Neural Networks and Signal Processing . 2003