From: Shallow and deep feature fusion for digital audio tampering detection
Methods
Classical dataset
GAUDI-DI dataset
Accuracy (%)
F1-score (%)
Nicolalde et al. [24]
92.08
92.54
83.53
78.84
Nicolalde et al. [23]
94.06
95.44
83.67
78.90
Esquef et al. [22]
83.17
82.11
79.02
74.18
Wang et al. [16]
95.05
95.41
84.99
79.14
Our method
97.03
96.91
88.31
88.17
