Feature Extraction Method for Hidden Information in Audio Streams Based on HM-EMD

Table 2

Fake audio detection results based on ASVSpoof 2019 dataset.

Classifier	Attack type	MFCC (EER %)	(EER %)	(EER %)
TridentResNet	LA	7.06	2.82	1.45
	PA	9.55	8.51	7.14
3L neural network	LA	9.95	4.98	3.84
	PA	11.83	11.79	9.44