From: Three-dimensional convolutional restricted Boltzmann machine for human behavior recognition from RGB-D video
Action
3DCRBM
Paper [30]
Walk
93.5%
96.5%
Sit
100%
91.5%
Stand
89%
Pick up
97.5%
Carry
71.9%
Throw
95%
59%
Push
92%
81.5%
Pull
92.5%
Wave
Clap
93%
Overall
92.9%
90.9%