Created by: yuzhimu

I want to change speech features using mfcc, but met a problem。How to change the input layer to solve this problem