可以在语音识别中添加过滤器程序吗? [英] can we add filter program in speech recognition?

问题描述

实际上，当我通过麦克风输入时，它无法正确反应.除了沉默以外，它还把所有东西都作为输入，因此是否可以在语音识别中添加过滤器程序?

actually when i give input through microphone it cannot react properly . other than silence it take everything as input so is it possible to add filter program in speech recognition? and if yes then how?

推荐答案

最简单的方法是查找阈值:如果音量低于该阈值，则将零发送到语音识别引擎.
更复杂的是噪声过滤.您必须对原始音频数据进行快速傅立叶变换，将它们与快速傅立叶变换后的噪声数据进行比较，然后删除/减少相应的频率.
实际上，一个好的降噪麦克风可以提供很多帮助.

The simplest way is to look for a threshold value: if the volume is below that value, send zeros to your speech recognition engine.
More complicated is noise filtering. You''ll have to do a Fast Fourier transformation of original audio data, compare them with Fast Fourier transformed noise data, and remove/reduce corresponding frequencies.
Actually, a good noise-cancelling microphone can help a lot.

这篇关于可以在语音识别中添加过滤器程序吗?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！