transfer

中国科学技术大学本科课程语音信号处理实验代码(speech processing code for USTCers)

绘制一段语音信号的波形并观察其频谱，用三角窗设计一个满足给定相应技术指标的FIR滤波器，对该语音信号进行滤波去噪处理，比较滤波前后的波形和频谱并进行分析。

针对信号自适应chirplet分解未知参数多、实现起来比较困难的特点，文献[1]提出了一种新的chirplet分解快速算法。该算法利用计算信号的二次相位函数，得到其能量分布集中于信号的调频率曲线上的结论，此时通过谱峰检测可同时获得chirplet调频率、时间中心和幅度的估计；然后通过解线性调频技术获得其初始频率和宽度的估计，仿真结果验证了本文算法的有效性。(this code includs one method for chirplet analasis,that is used in voice and ultra-wave analasys,etc.)

这是一个在线签名的程序，使用了AR实现算法和DTW算法(This is an online signature process, using the AR algorithm and DTW algorithm)

code for voice analysis final code for sapeech recognition

通常语音信号在增强时会出现混响现象，演讲者为了消除背景混响，不得不频繁地偏转头部的方向，这样会造成脉冲响应的不断改变。我们结合盲解卷法和频谱消去法来提高逆滤波器的滤波效果。我们利用输入语音信号间的相关系数矩阵计算出稳定、精确的室内脉冲响应的逆滤波器，而这些输入信号无需测量室内的脉冲响应就能被观测到。逆滤波能够消除早期的反射，这些反射包含混响中的绝大部分能量。之后，用频谱消去法来抑制逆滤波后的信号的尾部混响。本方法在实际适应性方面的表现通过具体的实验进行了验证，结果表明盲解卷法和频谱消去法的结合相较于单独使用一种方法，能够提供一个更优越的演说环境。(Usually voice signal will be enhanced when the reverberation phenomenon, the speaker in order to eliminate the background reverberation, and had to frequently deflect the direction of the head, this will cause the impulse response of changing. We combine the blind deconvolution and spectral elimination method to improve the filtering effect of the inverse filter. We use the correlation coefficient matrix between the input speech signal to calculate the stable, accurate inverse filter of the indoor impulse response, and these input signals do not need to measure the room impulse response can be observed. Inverse filtering to eliminate early reflections, these reflections contains most of the energy in the reverberation. After the spectrum elimination method to suppress the tail reverberation of the signal after the inverse filtering. Performance in the actual adaptation of the method by specific experimental validation results show that the blind deconvolution and spectrum to eliminate )

speaker recognition using MFCC GMM EM

语音识别，实现语音的识别功能，，特定人的语音识别。识别0~9的发音(Speech recognition, voice recognition function, a particular person' s speech recognition. Identify the pronunciation of 0 to 9)

文中研究了6种常用数字调制信号识别的特征参数集，并采用决策树判别方法进行分类识别。仿真结果表明，在SNR≥5dB时，识别正确率在99 以上，且当SNR≥20dB时，识别正确率达到100 。其特点是，算法简单，识别正确率高，达到了自动分类识别的目的，并有利于实现识别的实时化。(In this paper, we study the set of characteristic parameters of the six kinds of commonly used digital modulation signal recognition, and decision tree method for classification. The simulation results show that SNR 鈮� 5dB, the correct rate more than 99 , and when SNR 鈮� 20dB, the correct rate of 100 . Which is characterized by simple algorithm to identify the correct rate, to achieve the purpose of automatic classification and recognition, and help to identify real-time.)

Use of repeater transponder controls, WebForm controls