Auditory Feature Extraction for Speech Recognition