Anomaly Recognition Algorithm for Human Multipose Motion Behavior Using Generative Adversarial Network

Table 2

Abnormal recognition results.

Motion behavior	Total number of images (piece)	Actual abnormal quantity (piece)	Abnormal recognition quantity (piece)
Walking	351	38	38
Running	254	48	47
Kicking	184	52	51
Jumping	645	34	34
Standing	284	19	19
Squatting	168	27	27
Hands up	294	34	34
Reverse	587	29	28
Head down	667	47	46
Head up	566	63	62