Using librosa to get MFCC feature
y, sr = librosa.load(\'Test.wav\') mfcc = librosa.feature.mfcc(y=y, sr=sr) #result matriks (20 x t) np.shape(mfcc) # print(