Empirical comparison of analog and digital auditory preprocessing for automatic speech recognition