EURASIP Journal on Image and Video Processing

Table 10 Recognition results on KTH dataset

From: A framework of human detection and action recognition based on uniform segmentation and combination of Euclidean distance and joint entropy-based features selection

Method	Algorithm		Measures
Method	FS	WFS	PPV (%)	AUC	FNR (%)	RR (%)
W-KNN	✓		96.30	0.998	2.70	97.30
W-KNN		✓	82.70	0.961	15.80	84.20
SDA	✓		82.90	0.951	15.70	84.30
SDA		✓	70.78	0.947	23.50	76.50
LR	✓		77.30	0.945	21.0	79.0
LR		✓	68.81	0.928	25.20	74.80
M-class SVM	✓		98.93	0.993	0.7	99.30
M-class SVM		✓	92.1	0.978	6.70	93.30

Back to article page