The U.S. Patent and Trademark Office has granted NovoSpeech, a patent for one of its key speech recognition innovations, the virtual microphone array application, a proprietary algorithm that creates a number of independent channels —producing an array effect — from a single speech signal from one microphone.
Just as an array of microphones allows for signal enhancement in a noisy environment, NovoSpeech’s virtual microphone array enhances the speech signal against background noise, all without de-noising the speech signal, which often results in losing important data. Thus, NovoSpeech improves the robustness of speech recognition in real-life environments such as busy public places and moving vehicles and addresses one of the key challenges still facing speech recognition as it tries to penetrate mass markets.
NovoSpeech, and Israeli company, recently applied on an accelerated track for a U.S. patent rfor technology that lets users carry out very large vocabulary recognition in the small, constrained environments of mobile devices, according to company CEO Tina Ornstein.
"This is an incredible milestone for NovoSpeech and validates their cutting-edge technology. The ability to significantly enhance automatic speech recognition allows NovoSpeech a larger stake in a relatively untapped market. We look forward to the company's next developments," said Steve Rhodes, chairman and CEO of Trendlines, the parent company of NovoSpeech, in a statement.