中文摘要：

基于人耳听觉特性提出一种新的抗噪音识别特征：加权组合过零峰值幅度特征，是对过零峰值幅度特征的一种改进。加权组合过零峰值幅度特征以语音数据和差分语音数据作为处理对象，通过计算它们的上升过零率获得频率信息，经幅度非线性压缩获得密度信息，并根据人耳对声音的感知特点对其进行加权，形成最终的输出特征，识别网络使用HMM。仿真实现了使用新特征与原特征的算法识别结果，证明了新特征具有较高的识别率和优良的抗噪性能。

英文摘要：

This paper presents a new approach to extract anti-noisy speech feature： weighting combinition zero-crossings with peak amplitudes, which is based on human auditory model. It is an improved model of ZCPA（Zero-Crossings with Peak Amplitudes）. This approach uses the speech signal and it＇s difference signal as input. The frequency information of speech signal is obtained by upward-going zero-crossing intervals, and the intension information is incorporated by compressed nonlinear amplitudes. The speech feature is weighted based on the auditory characteristics by using weighting function and then the finally feature is outputed. The recognition part uses HMM. Experimental results of speech recognition demonstrate that this new feature is more robust than the older feature in noise environment.