Table 1: Recognition accuracies (unit: %) comparison on the MSRAction3D dataset and MSRGesture3D dataset.

MethodAccu. (%) on two datasets
MSRAction3DMSRGesture3D

DMM-HOG [3]85.589.2
Random Occupancy [22]86.588.5
Actionlet Ensemble [23]88.288.2
Depth Cuboid [24]89.390.5
Vemulapalli [25]89.587.7
DMM + CRC [26]92.392.5
DMM + ICRC (proposed)93.894.8