| 改进的基于增量估计的快速高斯计算 |
| |
| 引用本文： | 钱胜,吕萍,吴及.改进的基于增量估计的快速高斯计算[J].清华大学学报(自然科学版),2009(Z1). |
| |
| 作者姓名： | 钱胜 吕萍 吴及 |
| |
| 作者单位： | 清华-讯飞语音技术联合实验室;清华大学电子工程系; |
| |
| 摘 要： | 该文分析讨论了连续语音识别系统中的快速高斯计算问题。语音信号的短时平稳特性,使得相邻语音帧可能共享相似的分布。最大概率增量估计算法利用该特性,估计当前帧与基准帧间似然值增量的最大值,以减少似然值的精确计算量。该文针对该算法中增量上界被高估的问题,在增量上界平滑、最优G auss候选、风险因子设定等方面进行了改进。实验结果表明,在几乎不损失识别率的情况下,改进后的M P IE算法可节约40%的维数计算,解码速度相对提高10%。
|
| 关 键 词： | 语音识别 快速Gauss计算 |
Maximum probability increase estimation method for fast Gaussian likelihood computations |
| |
| QIAN Sheng,LU Ping,WU Ji.Maximum probability increase estimation method for fast Gaussian likelihood computations[J].Journal of Tsinghua University(Science and Technology),2009(Z1). |
| |
| Authors: | QIAN Sheng LU Ping WU Ji |
| |
| Institution: | 1.Tsinghua-iFlyTek Joint Laboratory for Speech Technologies;Beijing 100084;China;2.Department of Electronic Engineering;Tsinghua University;China |
| |
| Abstract: | This paper presents a fast Gaussian likelihood computational method for a continuous speech recognition system.Adjacent speech frames likely share similar distributions due to the semi-stationary feature of speech signals.Therefore,the maximum probability increase estimation(MPIE) algorithm can estimate the maximum probability increase between the current frame and a reference frame to reduce the explicit Gaussian likelihood computations.Overestimates of the maximum probability increase are reduced by incre... |
| |
| Keywords: | speech recognition fast Gaussian likelihood computation |
| 本文献已被 CNKI 等数据库收录！ |
