Design and Realization of FPGA-Based Perceptron
ANN (Artificial Neural Network) is a common learning method aiming at imitating human brain. At present, ANNs function mainly is simulated by the traditional computer. It is said that the parallel operating Neural Network is accustomed to be realized on the traditional operating computer with many kinds of application software. The applications of ANN-model urgently require hardware implementation of ANN system, instead of the traditional computer simulation. This paper presents an effective design and realization of FPGA-Based perceptron. Firstly, the model and its learning algorithm of single cell (perceptron) are introduced, then, the implementation of single cell on FPGA (Field Programmable Gate Array) is systemic described through an example in that the perceptron is applied to identify numbers. The experimental results indicate that the design in this paper is valid. The work in this paper is a primary exploration for the ANN hardware implementation and establishes the foundation for the further research.
Jihong Liu Baorui Li
College of Information Science and Engineering Northeastern University Shenyang, 110004;Brain Scienc College of Information Science and Engineering Northeastern University Shenyang, 110004
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）