1

I am developing an app for "speech to text", I am using openears for this. I am aware of its language model and dic files.

but is it a good idea to create language model of whole vocabulary? (Language model of english language is 125 MB in size)

How can I implement "speech to text" for many languages. Should I create language model for each language? and what about my build size if I import 10 language model in my app?

Thanks in advance.

Mayank Jain
  • 5,663
  • 7
  • 32
  • 65

1 Answers1

2

OpenEars developer here. OpenEars supports two languages and it doesn't perform large vocabulary recognition. This is described in its docs, so check them out for the answers to these questions and more.

Halle
  • 3,584
  • 1
  • 37
  • 53
  • can I have link of official doc? – Mayank Jain Jun 18 '14 at 11:20
  • The docs are in the OpenEars download (look inside the folder called Documentation – they are available in HTML, docset, epub and pdf format) and are also all on http://www.politepix.com/openears along with some supplementary info. – Halle Jun 18 '14 at 11:26