Does Kaldi have an API? Is it possible to use Kaldi only for audio transcription?
I guess I'm a little confused how Kaldi is organized. When I walk through the run.sh scripts, it is my current understanding that this starts the process from the very beginning by building an AM and LM around a pre-existing corpus. What If I already have an AM and LM that I would like to use with Kaldi? Is there currently a tutorial that focuses more on this 'plug-and-play' framework with out worrying too much about the details of the speech recognition process?
On a slightly related note, are you aware of any scripts that can convert CMU's Sphinx4 Acoustic models into the format Kaldi accepts?
Thanks,
- Jeff
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
At the moment Kaldi is targeted more at people who are building ASR systems
than those who just want to use them. We may in future cater to needs such
as yours.
Regarding API, the idea is that you would write your decoder code in C++,
patterned after code already in Kaldi.
Dan
Does Kaldi have an API? Is it possible to use Kaldi only for audio
transcription?
I guess I'm a little confused how Kaldi is organized. When I walk through
the run.sh scripts, it is my current understanding that this starts the
process from the very beginning by building an AM and LM around a
pre-existing corpus. What If I already have an AM and LM that I would like
to use with Kaldi? Is there currently a tutorial that focuses more on this
'plug-and-play' framework with out worrying too much about the details of
the speech recognition process?
On a slightly related note, are you aware of any scripts that can convert
CMU's Sphinx4 Acoustic models into the format Kaldi accepts?
Hi,
Does Kaldi have an API? Is it possible to use Kaldi only for audio transcription?
I guess I'm a little confused how Kaldi is organized. When I walk through the run.sh scripts, it is my current understanding that this starts the process from the very beginning by building an AM and LM around a pre-existing corpus. What If I already have an AM and LM that I would like to use with Kaldi? Is there currently a tutorial that focuses more on this 'plug-and-play' framework with out worrying too much about the details of the speech recognition process?
On a slightly related note, are you aware of any scripts that can convert CMU's Sphinx4 Acoustic models into the format Kaldi accepts?
Thanks,
- Jeff
At the moment Kaldi is targeted more at people who are building ASR systems
than those who just want to use them. We may in future cater to needs such
as yours.
Regarding API, the idea is that you would write your decoder code in C++,
patterned after code already in Kaldi.
Dan
On Thu, Feb 6, 2014 at 2:44 PM, Jeff Acquaviva jacquaviva@users.sf.netwrote: