论文信息 - Artificial Neural Network Optimization for FPGA

Artificial Neural Network Optimization for FPGA

This paper describes a cost effective artificial neural network implementation on an FPGA in three easy steps. Furthermore, it proposes the manner in which network layers are mapped into a particular hardware structure such that the performance and efficiency, with which the hardware resources are used, are greatly improved. A reconfigurable, parameterised neural node is presented as the basic building block for neural implementations, and is modelled in Verilog (HDL). The results show a high degree of parallelism, fast performance and most important low area resources.

Ivan Grech | Joseph Micallef | Edward Gatt | Mark Bonnici

[1] Nadia Nedjah,et al. Stochastic reconfigurable hardware for neural networks , 2003, Euromicro Symposium on Digital System Design, 2003. Proceedings..

[2] S. Y. Kung,et al. An algebraic projection analysis for optimal hidden units size and learning rates in back-propagation learning , 1988, IEEE 1988 International Conference on Neural Networks.

[3] Jihan Zhu,et al. FPGA Implementations of Neural Networks - A Survey of a Decade of Progress , 2003, FPL.

[4] Wenbin Liu,et al. A Posteriori Error Estimators for a Class of Variational Inequalities , 2000, J. Sci. Comput..

[5] Paramasivan Saratchandran,et al. Parallel architectures for artificial neural networks - paradigms and implementations , 1998 .

[6] Francisco José Ballester-Merelo,et al. Artificial neural network implementation on a single FPGA of a pipelined on-line backpropagation , 2000, Proceedings 13th International Symposium on System Synthesis.

[7] Tomas Nordström,et al. Highly parallel computers for artificial neural networks , 1995 .

[8] David Myers,et al. Backpropagation in linear arrays-a performance analysis and optimization , 1995, IEEE Trans. Neural Networks.

[9] Clark S. Lindsey,et al. Survey of neural network hardware , 1995, SPIE Defense + Commercial Sensing.