I know php well and I use javascript and jquery but I don't seem to know how to make a speech to text conversion with them though, but i do know that there are many flash speech recognition api's around but I would like a faster, I would like a script for this that can accurately use your voice and convert it into text. Thank you very Much, Anonymous.
I know php well and I use javascript and jquery but I don't seem to know how to make a speech to text conversion with them though, but i do know that there are many flash speech recognition api's around but I would like a faster, I would like a script for this that can accurately use your voice and convert it into text. Thank you very Much, Anonymous.
Share Improve this question edited Jan 1, 2012 at 20:27 Charles 51.5k13 gold badges106 silver badges144 bronze badges asked Jan 1, 2012 at 14:50 user1125155user1125155 411 gold badge1 silver badge1 bronze badge 3- I'd assume that one is impossible. I don't know any library to analyse sounds for PHP. – TimWolla Commented Jan 1, 2012 at 14:53
- 5 I wouldn't go so far as to say its impossible, just that any sane developer wouldn't choose PHP or JavaScript as candidates for doing speech recognition – jprofitt Commented Jan 1, 2012 at 14:55
- A JS speech recognition program may well be impossible. – benesch Commented Jan 1, 2012 at 15:52
3 Answers
Reset to default 2If your goal is to do speech recognition from an html page, you might want to look at some other alternatives. Chrome supports speech recognition for text input. See http://slides.html5rocks./#speech-input and http://www.filosophy/2011/03/talking-to-the-web-the-basics-of-html5-speech-input/.
These use the following tag for speech recognition:
<input type=”text” speech x-webkit-speech />
I believe Chrome is the only browser that currently supports this. http://tomlerendu./tutorial/how-to-use-html-5-speech-input/ has a good example and shows
if( document.createElement('input').webkitSpeech==undefined )
{
//no speech support
}
as a means to test if speech recognition is supported.
Historically, there have been other approaches. Opera implemented a different solution, but it appears they are no longer supporting it - http://dev.opera./articles/view/getting-to-know-voice/.
Another approach that has been used is to use a java applet or flash app that municates with a speech recognition back end. WAMI is a good example of this - http://wami.csail.mit.edu/. These approaches use a rich client (Java or Flash or other plug in) to capture speech and send it to a server or some local speech engine for processing.
In html 5 trere is a tag you can do it with that..
Read: link1 link2
You could have a look at speechapi. and build yourself what you need but if its the other way around;text to speech,speak.js will do.
Hope that helps.