| 基于MI_OneStage算法的连续数字语音识别 |
| |
| 引用本文： | 徐华,俞一彪. 基于MI_OneStage算法的连续数字语音识别[J]. 通信技术, 2003, 0(3): 1-3 |
| |
| 作者姓名： | 徐华 俞一彪 |
| |
| 作者单位： | 苏州大学电子信息学院,苏州,215021 |
| |
| 摘 要： | 语音识别的互信息匹配模型是依据互信息理论而提出的。在互信息匹配理论的现有成果的基础上,提出了基于互信息估计的连续语音识别算法—MI_OneStage算法,实验表明所提出的算法简单有效,优于传统的OneStage算法。
|
| 关 键 词： | 语音识别 互信息估计 MI_OneStage算法 |
| 修稿时间： | 2002-09-19 |
Connected Digit Speech Recognition Based on MI_OneStsge Algorithm |
| |
| Xu Hua Yu Yibiao. Connected Digit Speech Recognition Based on MI_OneStsge Algorithm[J]. Communications Technology, 2003, 0(3): 1-3 |
| |
| Authors: | Xu Hua Yu Yibiao |
| |
| Abstract: | Based on the existing fruit of Multi -Information Match theory,the author proposed the MI_OneStage algo-rithm ,which is applied on the connected digit speech recognition.The pr oposed algorithm is effective which is proved by the re-sult of the experiment. |
| |
| Keywords: | speech recognition multi -informa tion estimate MI_OneStage algorit hm |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
