Multimodal Semantics Extraction from User-Generated Videos

Table 6

Experimental results on identifying the optimal cameras. P stands for Precision, R for Recall, and F for F-measure.

Event	All cameras (index)	Ground truth sets of optimal cameras	Automatically identified optimal set of cameras	P	R	F
Football match 1	[1,2,3,4,5,6]	[1,2,3],[2,3,4,5],[5,6]	[2,3,4,5,6]	0.8	1.0	0.89
Football match 2	[1,2,3,4,5,6]	[1,2,3,4],[4,5,6]	[1,2,3,4]	1.0	1.0	1.0
Football match 3	[1,2,3,4,5]	[1,2],[2,3],[4,5]	[1,2,3]	0.75	1.0	0.86
Concert 1	[1,2,3,4,5,6,7,8,9]	[1,2,3,4,5,6,7,8,9]	[1,2,3,4,5,6,7,8,9]	1.0	1.0	1.0
Concert 2	[1,2,3,4,5,6,7,8,]	[1,2,3,4,5,6,7,8]	[1,2,3,4,5,6,7,8]	1.0	1.0	1.0
Average over all events	—	—	—	0.91	1.0	0.95