Speaker Recognition Based on Probabilistic Neural Network

摘要：

Probabilistic Neural Network(PNN) is considered a good speaker recognition method, but, the recognition performance of PNN is greatly depends on the selection of smoothing parameter. In this paper a new method is proposed to estimate the arrange of smoothing parameter based on the samples distances among different classes. The experimental results of speaker recognition based on PNN show that this method is effective.

关键词： Probabilistic Neural Network(PNN) Smoothing Parameter Speaker recognition

作者: Jinming Wang Geng Wang

作者单位: Institute of Communications Engineering, PLA Univ. of Sci. & Tech. Nanjing, China

会议类型: 国际会议

会议名称: 2011 4th International Congress on Image and Signal Processing(第四届图像与信号处理国际学术会议 CISP 2011)

会议地点: 上海

会议语种:英文

页码: 2468-2471

在线出版日期: 2011-10-15（万方平台首次上网日期，不代表论文的发表时间）