Text this: Auditory-Spectrum Quantization Based Speech Recognition