Peter Glen via Crowdrise
April 23, 2012
I am a software developer, started a voice recognition research project two years ago. The research creates a new method of voice recognition, that will make voice interface available on today's computers and handhelds.
As it is well known, state of the art in voice recognition is still in its infancy. The current systems have to be trained, and even after extensive training they still get a simple "yes / no" answer wrong. So I set out to remedy that. I observed, a dog can get several tens of words right, so a computer with billions of operations per second should be able to do better.
Any help would be appreciated,