VoxForge
VoxForge
The amount of speech audio required to improve the VoxForge Speaker Independent Acoustic Model for your voice depends on many things, however, the HTK manual says that a performance improvement should be observable with just 30 seconds of speech (around 20 utterances).
|
Note: the speech audio files you recorded in the How-to and Tutorial were recorded with a 48 kHz sampling rate at 16 bits per sample. The VoxForge Speaker Independent Acoustic Models we will use in this tutorial were trained with audio recorded at 8kHz:16-bit (we also have 16kHz:16-bits Acoustic Models - see the Nightly Builds directory on the VoxForge Repository). Because
of this,
you need to 'downsample' your speech audio files to 8kHz:16-bits and
then convert them to HTK '.mfc' files. For details on Acoustic Model Creation see this link. |
1. First, create a new directory called 'adapt' in your voxforge directory.
2. create a directory called 'wav' in your adapt directory.
3. create a file called 
FilesToBeDownsampled in your 'adapt' directory.
4.
copy the
downsample.pl
script to your adapt directory (note that if you download this file,
you need to rename it to 'downsample.pl' - otherwise it will download
as 'downsample_pl.txt').
Note: the downsample.pl script requires the the SoX sound processing utililiy. To confirm that it is included in your Linux distribution, type 'sox' at a command line, and see if you get an error saying "Failed: Not enough input or output filenames specified". If SoX is not included in your distro, use Yum to install it using the following commands: |
5. run the downsample.pl script as follows:
| $./downsample.pl FilesToBeDownsampled wav 48000 8000 |
6. Download the most current VoxForge Acoustic Model from the Nightly Builds directory, and put the following files in the new 'adapt' directory you just created: