10 Replies - 395 Views - Last Post: 02 September 2013 - 01:08 PM Rate Topic: -----

#1 fligex  Icon User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 6
  • Joined: 01-September 13

developing with speech-to-text

Posted 01 September 2013 - 04:47 PM

Hi everyone,

I want to begin writing a program that utilizes some type of speech-to-text engine but I am afraid that I do not quite know where to begin. I know that at some point I need to import the text result into a c++ program that I will write. It would probably be best that I don't use some other program to accomplish this. I prefer to be able to invoke a library function, or three, of some kind within my own program that will retrieve the microphone input and return the text result so that I can apply some operations to the text it provides. I hope I am explaining myself clearly if not please ask me questions to help clarify further. Any information anyone could offer in how to accomplish this or how to go about it would be welcomed. Thanks.

Is This A Good Question/Topic? 0
  • +

Replies To: developing with speech-to-text

#2 Skydiver  Icon User is offline

  • Code herder
  • member icon

Reputation: 3172
  • View blog
  • Posts: 9,611
  • Joined: 05-May 12

Re: developing with speech-to-text

Posted 01 September 2013 - 05:44 PM

What OS are you trying to do this on? Does it need to be realtime speech to text, or can it be done offline?
Was This Post Helpful? 0
  • +
  • -

#3 snoopy11  Icon User is offline

  • Engineering ● Software
  • member icon

Reputation: 710
  • View blog
  • Posts: 2,033
  • Joined: 20-March 10

Re: developing with speech-to-text

Posted 01 September 2013 - 05:48 PM

Hi,

If This is Windows

There is Microsofts Speech Api... or sapi.

you simply #include <sapi> and initialise COM

You take a pointer to an ISPVoice

ISPVoice* pVoice;


Use CoCreateInstance with the CLSID set to CLSID_SpVoice

use the Voice then when finished free the Voice with the Release method

Use CoUnInitialise to end.

Regards

Snoopy.

This post has been edited by snoopy11: 01 September 2013 - 05:49 PM

Was This Post Helpful? 1
  • +
  • -

#4 fligex  Icon User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 6
  • Joined: 01-September 13

Re: developing with speech-to-text

Posted 01 September 2013 - 05:53 PM

Well I am currently planning on doing this on Linux (Ubuntu).

I'm not sure I know enough about it to conclude one over the other but I would take a guess and say that realtime would be needed.
Was This Post Helpful? 0
  • +
  • -

#5 fligex  Icon User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 6
  • Joined: 01-September 13

Re: developing with speech-to-text

Posted 01 September 2013 - 05:59 PM

View Postsnoopy11, on 01 September 2013 - 05:48 PM, said:

Hi,

If This is Windows

There is Microsofts Speech Api... or sapi.

you simply #include <sapi> and initialise COM

You take a pointer to an ISPVoice

ISPVoice* pVoice;


Use CoCreateInstance with the CLSID set to CLSID_SpVoice

use the Voice then when finished free the Voice with the Release method

Use CoUnInitialise to end.

Regards

Snoopy.


Thank you Snoopy I will look into the speech API from Microsoft. I haven't done any c++ programming on a windows machine yet though so i'm not sure how to even begin with that either. VS I would assume could help me out there.
Was This Post Helpful? 0
  • +
  • -

#6 fligex  Icon User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 6
  • Joined: 01-September 13

Re: developing with speech-to-text

Posted 01 September 2013 - 06:19 PM

Is there a linux equivalent to the MS speech API? I would be interested in information regarding that as well. I'm in the phase of absorbing as much information regarding this topic as possible so feel free to keep the suggestions coming. In the meantime I'll try to find as much info as I can on the MS SAPI. Can anyone offer reasoning as to why one or other would be more beneficial for speech-to-text? (linux vs Windows) Are there any other suggestions or notes pertaining to this subject that I should know? Thanks all.
Was This Post Helpful? 0
  • +
  • -

#7 Adak  Icon User is offline

  • D.I.C Lover
  • member icon

Reputation: 331
  • View blog
  • Posts: 1,168
  • Joined: 01-April 11

Re: developing with speech-to-text

Posted 01 September 2013 - 06:30 PM

Three suggestions:

1) If you want to use Ubuntu, be SURE to repeat your question in the very active Ubuntu forum:

http://ubuntuforums.org/index.php

2) Ubuntu as you may know, is based on Debian, which also has a large user base, so check with them.

3) Google it to see what else might be available from other Linux distro's.
Was This Post Helpful? 0
  • +
  • -

#8 snoopy11  Icon User is offline

  • Engineering ● Software
  • member icon

Reputation: 710
  • View blog
  • Posts: 2,033
  • Joined: 20-March 10

Re: developing with speech-to-text

Posted 01 September 2013 - 06:31 PM

Hi,

for linux there is the Java based api

FreeTTS

Regards

Snoopy.
Was This Post Helpful? 0
  • +
  • -

#9 fligex  Icon User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 6
  • Joined: 01-September 13

Re: developing with speech-to-text

Posted 01 September 2013 - 07:00 PM

Thank you for your suggestions, I will look into them all and hopefully return here with more questions. Thank you all for being so helpful.
Was This Post Helpful? 0
  • +
  • -

#10 Skydiver  Icon User is offline

  • Code herder
  • member icon

Reputation: 3172
  • View blog
  • Posts: 9,611
  • Joined: 05-May 12

Re: developing with speech-to-text

Posted 01 September 2013 - 09:06 PM

My read of the description of FreeTTS is that it converts text to speech, but the OP seems to be looking for speech to text.
Was This Post Helpful? 0
  • +
  • -

#11 fligex  Icon User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 6
  • Joined: 01-September 13

Re: developing with speech-to-text

Posted 02 September 2013 - 01:08 PM

View PostSkydiver, on 01 September 2013 - 09:06 PM, said:

My read of the description of FreeTTS is that it converts text to speech, but the OP seems to be looking for speech to text.


This is true Skydiver, I am primarly looking for speech-to-text. I will eventually want to use text-to-speech but it is not strictly required like speech-to-text is.
Was This Post Helpful? 0
  • +
  • -

Page 1 of 1