|
One of the most exciting areas for Brainhat is improved speech recognition.
Because Brainhat is knowledge-based, it can anticipate what it might
hear next, and use context to better improve recognition rates.
For those recognition systems that require a context-free grammar,
such as VoiceXML or telephony speech engines, Brainhat can generate
dynamic grammars.
At any time, one can ask Brainhat to list all the things it is expecting
to hear.
How would one use Brainhat's ability to anticipate input, generate dynamic grammar and track a conversation to improve speech recognition? One straightforward use would be as a "shim" between speech engine and application:
Brainhat, Inc. sees improvements in speech recognition through the application of knowledge as one of the key components for bringing the recognition business forward, and for enabling free-form dialog between humans and machines. Take a look at our technology. It could change the speech recognition world.
|