The paper introduces the definition and estimation methods of speaking rate in speech recognition, and presents a novel robust estimation method.
分析了语音识别系统中语速的定义与估计方法，提出了基于持续时间分布的鲁棒语速估计法。
Therefore a robust estimation of DOA can be achieved with this novel adaptive antenna.
因此基于该新型自适应天线，可以实现来波方向的鲁棒性估计。
Critical values of the equivalent weight function are often fixed in constants in robust estimation.
抗差估计等价权函数中临界值一般定为常数。
This paper discusses the problems of optimizing parameters and robust estimation for GPS height fitting.
本文研究了GPS高程拟合时对所建立的数学模型进行参数优化选择和稳健估计。
Through robust statistics, put forward a robust estimation algorithm to estimate measuring results and error.
通过鲁棒统计，提出了一个鲁棒估计算法处理测量结果和误差。
Results The results of robust estimation are better in the simulation and example analysis.
结果模拟及实例分析可见稳健估计的结果较好。
By simulation test, it shows that robust estimation, which is applied to parameter estimation, can resist gross errors.
通过使用模拟数据进一步验证了稳健估计用于参数估计中良好的抗差作用。
Conclusion it is necessary to carry out robust estimation to generalized additive models when there are outliers in data.
结论在离群点存在时对广义可加模型进行稳健估计是必要的。
The method got global motion parameters by local motion vectors through quick M-robust estimation; these parameters are filtered to remove stochastic motion noise.
采用新的快速M鲁棒估计法获得摄像机全局运动参数集；滤波该参数集滤除随机抖动带来的运动噪声。
The algorithm can not only track rapidly changed pitch, and enhance pitch estimation with irregular pulse, but also make robust estimation of sharply changed pitch.
这种算法不仅能够跟踪快变的基音周期，增强不规则脉冲时基音周期的估计；还能够鲁棒地估计陡变的基音周期。
The equivalent weight function in robust estimation is usually constructed by the statistics of normal distribution. The critical values are often fixed by experience.
抗差估计等价权函数一般由正态分布统计量构造，其临界值（或称准则）一般由实际经验确定。
A robust multiscale algorithm of hierarchical estimation for gradient-based motion estimation was proposed, by a combination of robust estimation method and multiscale technique.
将鲁棒估计方法与多尺度方法相结合，给出了一种基于梯度的鲁棒分层估计多尺度运动估计算法。
Based on the robust estimation in the state-space and the adaptive modification of prediction covariance matrix, a new algorithm was proposed to track Thevenin equivalent parameters.
融合预报协方差矩阵的自适应调整和状态空间鲁棒性估计，给出辨识戴维南等值参数的新算法。
Based on some prior knowledge, the proposed approaches utilize fewer parameters to describe the variation between speakers, and thus fewer adaptation data are needed to give robust estimation.
这两种方法在一定先验知识的基础上采用较少的参数来描述说话人间的差异，因而只需要少量的自适应数据就可以得到参数的鲁棒估计。
Zero-error experiment can observe the distribution of gross error after robust estimation, then compare the estimation value and real value, so we can evaluate externally all kinds of robust method.
零误差实验可以更有效地观察粗差在经过抗差估计以后的分布情况，再将估计值与真值比较，从而较客观地对各种抗差估计方法做出评价。
The results showed that the application of Robust Estimation Theory in forestry gray modeling could not only avoid the effect of gross error, but also reach the predicted precision over 80%.
结果表明，利用稳健估计理论来进行林业灰色建模，不但可以抗击粗差的影响，而且预测精度基本满足在80%以上。
The results showed that the application of Robust Estimation Theory in forestry gray modeling could not only avoid the effect of gross error, but also reach the predicted precision over 80%.
结果表明，利用稳健估计理论来进行林业灰色建模，不但可以抗击粗差的影响，而且预测精度基本满足在80%以上。
应用推荐