├── README.md
└── mfcc.py


/README.md:
--------------------------------------------------------------------------------
1 | # comparing-audio-files-python
2 | This project is for the comparison of two audio files based on their MFCC's.
3 | 
4 | 
5 | Please feel free to submit a PR with commented code (if needed).
6 | 


--------------------------------------------------------------------------------
/mfcc.py:
--------------------------------------------------------------------------------
 1 | import librosa
 2 | import matplotlib.pyplot as plt
 3 | from dtw import dtw
 4 | 
 5 | #Loading audio files
 6 | y1, sr1 = librosa.load('/home/user/Desktop/1.wav') 
 7 | y2, sr2 = librosa.load('/home/user/Desktop/2.wav') 
 8 | 
 9 | #Showing multiple plots using subplot
10 | plt.subplot(1, 2, 1) 
11 | mfcc1 = librosa.feature.mfcc(y1,sr1)   #Computing MFCC values
12 | librosa.display.specshow(mfcc1)
13 | 
14 | plt.subplot(1, 2, 2)
15 | mfcc2 = librosa.feature.mfcc(y2, sr2)
16 | librosa.display.specshow(mfcc2)
17 | 
18 | dist, cost, path = dtw(mfcc1.T, mfcc2.T)
19 | print("The normalized distance between the two : ",dist)   # 0 for similar audios 
20 | 
21 | plt.imshow(cost.T, origin='lower', cmap=plt.get_cmap('gray'), interpolation='nearest')
22 | plt.plot(path[0], path[1], 'w')   #creating plot for DTW
23 | 
24 | plt.show()  #To display the plots graphically
25 | 


--------------------------------------------------------------------------------