Table 5.

uulmMAD key pose versus all frame learning.

	Framewise		Majority voting
	Key poses	All frames	Key poses	All frames
Key poses	88.65 ± 5.66	67.56 ± 6.06	96.73 ± 2.84	93.29 ± 7.05
All frames	85.84 ± 7.15	72.84 ± 8.25	95.27 ± 3.93	94.70 ± 7.03

Bold values indicate the maximum average recall for framewise and full sequence majority voting classification schemes.