The goal of this project is to carry out fundamental research into the possibilities and difficulties of applying RC to some challenging auditory pattern recognition applications, such as recognizing spatial environment through multi-channel active or passive sonar signal and continuous speech recognition in background noise.