The Automatic Detection of Pedestrians under the High-Density Conditions by Deep Learning Techniques

Table 1

The information of training sets.

Name	Run	Number of images	Objects in each image	Total objects used
T1	16-1-1	110	25  110	2750
T2	16-1-1, 16-2-1	238	25  110, 55  128	9790
T3	16-1-1, 16-2-1,	543	25  110, 55  128,	23185
	17-1-1, 18-1-1		49  150, 39  155
T4	16-1-1, 16-2-1,	933	25  110, 55  128,	57295
	17-1-1, 18-1-1,		49  150, 39  155
	17-2-1, 18-2-1		99  210, 74  180