首页 > 代码库 > Ubuntu中安装编译并测试HTK语音识别库

Ubuntu中安装编译并测试HTK语音识别库

1.在网上看到首先必须确保电脑上安装了g++和libx11

g++ --version 
//检测g++版本
sudo apt-get install libx11-dev:i386

2.然后可从HTK官网 上 下载 HTK语音识别库

3.解压:sudo tar xzf HTK-3.4.1.tar.gz

4.打开文件夹:cd htk

5.在/usr/local/目录下安装文件:sudo ./configure --prefix=/usr/local/

出现:

 setting
config.status: creating HTKLVRec/Makefile
config.status: WARNING:  HTKLVRec/Makefile.inseems to ignore the --datarootdir setting
config.status: creating Makefile
config.status: WARNING:  Makefile.inseems to ignore the --datarootdir setting
**************************************************
 
HTK is now ready to be built.
 
Type"make all"to build the HTK libraries
and tools.
 
Then"make install"toinstallthem.
 
The tools will be installedin/usr/local//bin
 
Build notes: Language Modelling tools will be
built. HDecode will not be built. You can build
it manually later by running makehdecode
install-hdecode
**************************************************

然后,我按照上述步骤进行: make all

结果出现错误,找不到一个文件夹或者目录:

技术分享

 在网上搜了搜解决方案:

sudo apt-get purge libc6-dev
sudo apt-get install libc6-dev

sudo apt-get install libc6-dev-i386

解决~

 然后又报错.... 下载的语音库的问题,其中“空格键”与“Tab”键的混用,导致编译出错,这个语音库自动设置是不可改写的,所以我研究了半天页不太对。

在网上搜了搜,于是我打算下载KALDI语音库....

那就再写一篇说说KALDI语音库,无语...


Ubuntu中安装编译并测试HTK语音识别库