1

I'm writing an (offline) voice recognition app. I have CMU Sphinx4 set up and working using some of the included demo dictionaries. However, they're of limited scope (eg..numbers, cities, etc).

Is there a more comprehensive grammar available? Or maybe a repository of more of these limited grammars? I'm trying to exhaust any other options before creating my own.

Thank you

Ari Russo
  • 273
  • 2
  • 10

3 Answers3

2

Grammars are always specific to your particular goal, so it does not make sense to share those . Even such simple subject as digits can vary between concrete applications: we use "zero" and "oh" to denote "0" in regular speech, whilst scientists also use "not" for the same purpose.

Sphinx4 supports JSGF and GRXML formats, you can easily find specifications of both.

Alexander Solovets
  • 2,447
  • 15
  • 22
  • Thanks. Respectfully, I'll disagree - I think it does still make sense to share those for other people with the same goal. – Ari Russo Jan 04 '14 at 21:36
  • That said, my program is a kind of experimental passive conversation listener. So the comprehensive grammar IS suited to my goal. The HTML5 Web Speech API uses such a grammar but it's only available via web service. Thanks again – Ari Russo Jan 04 '14 at 21:37
  • 1
    Arn't you confusing grammar with dictionary or language model? In Sphinx4 a grammar is a set of rules describing what final hypothesis should look like. – Alexander Solovets Jan 04 '14 at 23:25
  • I may be -- my impression is that the grammar is the collection of dictionary and language models – Ari Russo Jan 04 '14 at 23:43
  • I believe using voxforge will accomplish what I am out to do – Ari Russo Jan 05 '14 at 04:09
0

You seem to be mistaking grammars with dictionaries. They are completely different things.

Sphinx supports not only grammars, but also n-gram language models. You may find them more versatile. Such model can be automatically generated and will work if given a large corpora which reflects the real usage sentences.

As for dictionaries - creating them for english is relatively simple. One could even think about a tool which reads a phonetic word representation from an online dictionary and converts it to sphinx format. The only input would be a word list.

Dariusz
  • 21,561
  • 9
  • 74
  • 114
0

I believe this paper will come handy to your effort. This paper entails creating grammar and dictionary for a new language, Swahili

Shadrack Kimutai
  • 143
  • 1
  • 7