| 基于矢量量化的仿生模式识别方法研究 |
| |
| 引用本文： | 殷建,殷业,徐毅.基于矢量量化的仿生模式识别方法研究[J].无锡职业技术学院学报,2007,6(5):31-34. |
| |
| 作者姓名： | 殷建 殷业 徐毅 |
| |
| 作者单位： | 1. 无锡职业技术学院,江苏,无锡,214121
2. 上海师范大学数理信息学院,上海,200234 |
| |
| 摘 要： | 该文探索了另一种新的仿生模式识别算法:基于矢量量化的仿生模式识别,用高维空间闭域作为同源类模板,并将其应用于语音关键词识别。实践表明仿生模式识别方法有许多优点,能够有效降低识别的错误率,识别效率优于传统模式识别方法。
|
| 关 键 词： | 语音识别 模式识别 仿生模式识别 矢量量化 高维流形 |
| 文章编号： | 1671-7880(2007)05-031-04 |
| 修稿时间： | 2007年10月15 |
Research on Bionic Pattern Recognition on Vector Quantization Method |
| |
| Authors: | YIN Jian YIN Ye XU Yi |
| |
| Abstract: | The paper researches another kind of new bionic pattern recognition algorithm: bionic pattern recognition of vector quantization method, using close fields of high-dimensional space as homology class templets, and applying it to speech keywords recognition. Practice indicates bionic pattern recognition method having many advantages, which can decrease error rate of recognition effectively. Its recognition efficiency excels traditional pattern recognition methods. |
| |
| Keywords: | Speech Recognition Pattern Recognition Bionic Pattern Recognition Vector Quantization High-Dimensional Manifold |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
