ASLNet: An Encoder-Decoder Architecture for Audio Splicing Detection and Localization

Table 4

The detection results of the Jadhav et al. [11] and the ASLNet with MFCC feature and  = 0.5 under the cross-database testing scenario.

Methods	Train dataset	Test dataset
Jadhav et al. [11]	ENSet2s	CNSet2s	0.8002	0.8512	0.8468
	CNSet2s	ENSet2s	0.8295	0.8395	0.8320
	ENSet3s	CNSet3s	0.8965	0.8620	0.8754
	CNSet3s	ENSet3s	0.8495	0.8929	0.8826
ASLNet	ENSet2s	CNSet2s	0.9072	0.9600	0.9420
	CNSet2s	ENSet2s	0.8990	0.8540	0.8718
	ENSet3s	CNSet3s	0.9347	0.9944	0.9740
	CNSet3s	ENSet3s	0.9474	0.9229	0.9271