Voice Recognition Based on Adaptive MFCC and Deep Learning

In this paper, we propose an enhanced voice recognition method using Adaptive MFCC and Deep Learning. To improve the voice recognition rate, it is important to extract the audio data from original signal. However the existing Algorithms which is used to remove the noise of particular band deteriorate the audio signal. Differently from the existing MFCC, the proposed filter is built up compactly in the data density area to reduce data loss, and impose the weighted value to the data area. As a result, it prevents the data loss which results in improving the recognition rate. In addition the Deep Learning makes it possible to use the Voice recognition without DB. Therefore, it can be effectively used for electronic devices with small memory.