Multiview Layer Fusion Model for Action Recognition Using RGBD Images

Table 7

Comparison of some action recognition approaches based on ease of use, perspective robustness, and real-world application conditions.

Approach	Flexible calibration	View scalability	Perspective robustness	Real-world Application Condition
				Complexity	Cumbersomeness from sensor/device attachment	Privacy
Inertial sensor-based [1–12]	N/A	N/A	High	Highly complex	Cumbersome	High
RGB vision-based [13, 17, 21–47, 71–73, 75]	Low-moderate	Low-moderate	Low-moderate	Simple-moderate	Not cumbersome	Lacking
Depth-based Single-view [14–16, 18–20, 70, 74]	Low-moderate	Low-moderate	Low-moderate	Depends	Not cumbersome	Moderate-high
RGB/depth-based multiview [48–68]	Low-moderate	Low-moderate	Low-moderate	Depends	Not cumbersome	Depends
Our work, depth-based multiview	Moderate-high	Moderate-high	High	Simple	Not cumbersome	High