Fig. 3From: Voice spoofing countermeasure for voice replay attacks using deep learningExtraction of the Decomposed Features from MFCCBack to article page