Human Activity Recognition Using Gaussian Mixture Hidden Conditional Random Fields

Table 9

Weighted average recognition rates of the proposed method with the existing state-of-the-art methods (unit: %).

State-of-the-art works	Average classification rates	Standard deviation
GMM	63.3	±2.7
SVM	67.5	±4.4
HMM	82.8	±3.8
Embedded HMM	85.9	±1.9
[62]	92.1	±3.2
[63]	84.3	±4.9
[18]	93.6	±2.7
[19]	93.0	±1.6
[22]	92.7	±2.5
[64]	80.1	±3.2
Proposed method	97.2	±2.8