A python scr ipt to recognise speech and convert to text.. implemented as a ROS node, which will publish the detected speech

该程序中audiorec.py可以录制一个音频文件，stft.py可以对该文件做短时傅里叶变化进行分析，得到音频文件的频率信息，进而可以根据音频文件的来源判断其中的信息。在压缩包中还包含一个音频文件可以作为例子。- The program audiorec.py can record an audio file, stft.py can make the file changes short time Fourier analysis to obtain frequency informat

使用python库里面的pydub，numpy来计算音频文件的总谐波失真(use python module pydub to calculate the wav thd)