How to use CMU Sphinx speech recognition with a Ruby application?

Question

I am looking for how to use CMU Sphinx with a Ruby (Rails) application. I need a very simple task - I have an mp3 file, and I want it to be transcribed into text.

How can I implement this easiest way? I do not know C / C ++, and my task is not so great as to learn C / C ++ :)

Thanks for the help!

Answer 1

CMUSphinx provides several interfaces that you can use to use speech recognition features. Some of them may be more suitable for you, and some less:

• Use command line tools and execute them as external binaries from a Rails application to get the desired results. The tool to execute is pocketsphinx_continuous. For more information on how to call binary files from Rails, see Question: How to execute a binary file on a hero?

• You can use the Sphinx4 framework with JRuby using the JVM, see for example how to use Sphinx4 from JRuby : http://cmusphinx.sourceforge.net/wiki/tutorialsphinx4#writing_scripts

• You can implement pocketsphinx bindings using SWIG . The easy part is that swig wrappers for Python already exist as part of pocketsphinx, you just need to use SWIG to create Ruby wrappers: https://sourceforge.net/p/cmusphinx/code/11643/tree/trunk/pocketsphinx/ swig /

• Finally, you can embed a REST website using Java REST frameworks to convert audio to text using CMUSphinx tools and invoke a selector from your Ruby code. For more information, see How To Use REST from Rails. This way your system will be truly scalable.

Answer 2

I recently created a pearl for this:

https://github.com/watsonbox/pocketsphinx-ruby

It uses the Ruby External Function Interface (FFI) to interact directly with libpocketsphinx.

Answer 3

Perhaps take a look at this Ruby-based web-based speech recognition service that uses PocketSphinx:

https://github.com/alumae/ruby-pocketsphinx-server