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

Table 1

Average results of acoustic streams classification based on DCASE dataset.

Classifier	MFCC
	(accuracy %)	(accuracy %)	(accuracy %)
TridentResNet_DevSet	73.7	74.5	75.0
TridentResNet_EvalSet	73.7	74.5	75.0
TridentResNet_Ensemble	74.2	75.0	75.5
TridentResNet_Weighted_Ensemble	74.4	75.2	75.7
3L neural network	54.1	60.8	71.5