Real-Time Source Separation

About

Improved speech recognition, intelligibility and noise reduction using real-time source separation This technology provides real-time separation of sounds from a mixture of sound sources using a very small array of 3 or 4 microphones. The technology can be used to improve speech recognition or intelligibility, to separate individual sound sources, and in a noisy environment for removing unwanted sounds. Status: There is a Granted US patent and a number of pending patents for this technology.