Multimodal Semantics Extraction from User-Generated Videos

Table 7

Experimental results for detecting in-view cameras. P stands for Precision, R for Recall, and F for F-measure.

Event	P	R	F
Football match 1	0.86	0.71	0.77
Football match 2	0.78	0.58	0.67
Football match 3	1.0	0.78	0.88
Concert 1	0.74	0.69	0.71
Concert 2	0.76	0.81	0.78
Average over all events	0.83	0.71	0.77