VoxForge
HTK's licence requires you to register before you can download the toolkit. The software is open source but there are limitations on the distribution of the HTK Toolkit itself. However, there is no limitation on the distribution of the models you create with the toolkit.
Create a new directory in your home directory called 'bin', it
should have the following path (replace 'yourusername' with the user name
you are using on your system):
click the following links:
and save them to your new bin directory.
Click the following link:
and save it to your new bin directory.
The HTK book is an excellent reference to the toolkit commands.
However, it is written for speech scientists and can be very
confusing to beginners.
Extract the files using:
this should create the following directories in your bin folder:
move the 'samples' and 'htkbook' directories to your htk-3.4 folder.
If you have a newer version of the gcc compiler (version 4 or above), you will need to install gcc version 3.4 compatibility modules so that HTK will compile properly. Use gcc's version command to see which version is installed on your system:
|
$gcc -v Target: x86_64-redhat-linuxConfigured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-bootstrap --enable-shared --enable-threads=posix --enable-checking=release --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-languages=c,c++,objc,obj-c++,java,fortran,ada --enable-java-awt=gtk --disable-dssi --enable-plugin --with-java-home=/usr/lib/jvm/java-1.5.0-gcj-1.5.0.0/jre --enable-libgcj-multifile --enable-java-maintainer-mode --with-ecj-jar=/usr/share/java/eclipse-ecj.jar --disable-libjava-multilib --with-cpu=generic --build=x86_64-redhat-linux Thread model: posix gcc version 4.3.0 20080428 (Red Hat 4.3.0-8) (GCC) |
If you have version 4.0 or above (I have version 4.3.0) use yum to install the required files to your system:
|
$su #yum install compat-gcc-34-c++ compat-gcc-34 |
After unpacking the sources, open a command line terminal and go to the /hom/yourusername/bin/htk3.4 directory where you downloaded your files.
configure
The default location for binaries is "/usr/local" which will put the tools in "/usr/local/bin". You need to change this default location using the "./configure" script to specify where you want the binaries installed:
$./configure --prefix=/home/yourusername/bin/htk-3.4
This directs the make command to put all your binaries in the following folder:
- /home/yourusername/bin/htk-3.4/bin.linux
Easiest way to do this is as follows (thanks to aaa in this thread):
# linux32 bash
↑ 32 bit terminal
#./configure --prefix=/home/yourname/htk
(may also need to: yum install libx11-devel)
#make all
#make install
Old approach:
After unpacking the sources, open a command line terminal and go to the /hom/yourusername/bin/htk3.4 directory where you downloaded your files.
You now need to install 386 compatibility versions of some development libraries:
$su
#yum install glibc-devel.i386 libX11-devel.i386
configure
The default location for binaries is "/usr/local" which will put the tools in "/usr/local/bin". You need to change this default location using the "./configure" script to specify where you want the binaries installed:
$./configure CC=gcc34 --prefix=/home/yourusername/bin/htk-3.4
This directs the make command to put all your binaries in the following folder:
- /home/yourusername/bin/htk-3.4/bin.linux
To build the libraries and binaries, execute the following:
|
$make all |
Running the following command will install them:
|
$make install |
Create a new directory called "HTK_scripts" in your voxforge folder, it should have the following path:
Copy
the following HTK scripts to your newly created 'HTK_scripts'
folder:
Julius has no limitations on distribution. It uses Acoustic Models in HTK format, and Grammar files in its own format.
Julius is a large vocabulary continuous speech recognition (LVCSR) engine. Julius is used in dictation applications.
Julian is a special version of Julius that performs grammar based speech recognition. Julian is used for command and control applications.
We will be using the Julian version of Julius for this tutorial.
click the following link:
and save it to your '/home/yourusename/bin' directory.
(note: if you want to compile Julius from source, see this FAQ entry)
Extract the file using:
this should create a julius-3.5.2 directory in your bin folder.
To update your user path, you need to add the following directories your user path variable:
To do this, edit your '.bash_profile' file in your home
directory (in Fedora you need to show 'hidden files' in Nautilus - so
you can display file names with a period in front of them). You do this
by adding the listed paths, separated by a colon (":") to the end of
the PATH variable as follows (all one line, no spaces):
|
# User specific environment and startup programs |
Log out and log back in to make your path change effective.
if your system lists all the options available to the hvite command, then HTK is installed properly.
if your system displays version information for Julius, then Julius is installed properly;
If you are using Linux (we will be using the Fedora for these tutorials) use the following command to download and install Audacity (as superuser):
$su |