Multimode Gesture Recognition Algorithm Based on Convolutional Long Short-Term Memory Network

Table 2

The accuracy results on the VIVA dataset.

Method	Accuracy
HOG + HOG2 [35]	65.51
CNN : LRN [15]	75.41
CNN : LRN : HRN [15]	78.51
C3D [14]	78.41
I3D [32]	84.11
MTUT [36]	87.09
3D-Dense (a)	89.22
Res3D + TCNs (b)	86.98
T3D-Dense + TCNs (c)	91.74
Proposed	92.55