Statistical class-based MFCC enhancement of filtered and band-limited speech for robust ASR | Publicación