General discussion

Locked

Voice Recognition Technology

By high-tech redneck ·
I have an application in mind that requires some voice recognition software/hardware and programming. From my first glance into the voice recognition software arena I got the feeling that it may not quite be there yet as far as being able to accurately capture billing type data. Does anyone have any experience/input about this technology that they would please pass on?

This conversation is currently closed to new comments.

5 total posts (Page 1 of 1)  
| Thread display: Collapse - | Expand +

All Comments

Collapse -

Voice Recognition Technology

by Brett Feldon In reply to Voice Recognition Technol ...

The technology probably is there, but it's at the stage where you'll need to define your requirements quite precisely to make a decision. There's currently a big trade-off between flexibility of input and the number of speakers which may be accurately recognised. For complete input freedom (free-text transcription) you need to look at products like ViaVoice, Dragon Dictate etc. For complete (or nearly complete) speaker freedom you need to look at Nuance, SpeechWorks, Philips etc.
I have some experience with Nuance systems. They work well, for what they're designed for. And that's the important point - you need to specify where the person is speaking, what medium the data needs to go through, how many different speakers need recognition, whether the possible things the speaker can say are fixed (or at least expressible with the kind of grammar you might write with yacc), and so on. Once you've defined it clearly, you can start narrowing down the products, until hopefully you find one which matches your requirements.

Good luck.

Collapse -

Voice Recognition Technology

by high-tech redneck In reply to Voice Recognition Technol ...

Thanks again Brett!

Collapse -

Voice Recognition Technology

by Mohammed Badran In reply to Voice Recognition Technol ...

Doing a voice recognaiztion project needs some math stuff to change your anlog i/p from the mic into a o/p digital signals that can be done by Fourier Transform and by takeing big number of samples you may accuretly pick the voice you want correctly

Collapse -

Voice Recognition Technology

by high-tech redneck In reply to Voice Recognition Technol ...

Poster rated this answer

Collapse -

Voice Recognition Technology

by high-tech redneck In reply to Voice Recognition Technol ...

This question was closed by the author

Back to Web Development Forum
5 total posts (Page 1 of 1)  

Software Forums