Visual Tracking Based on an Improved Online Multiple Instance Learning Algorithm

Table 1

The average center location errors for all of the trackers. Bold indicates the best performance.

Video clip	MIL	WMIL	Significance-coefficients MIL	IMIL
David indoor	23.631	19.435	17.716	14.739
Occluded face 2	33.805	22.165	15.747	15.368
Coke can	37.158	8.720	8.777	9.905
Cliff bar	43.994	18.225	18.882	11.97
Tiger 2	88.851	17.844	17.279	13.972
Coupon book	26.960	22.232	14.774	7.179