Research Article

Mining Key Skeleton Poses with Latent SVM for Action Recognition

Table 6

Results on Florence 3D Actions dataset.

Subject 12345678910Average

Wave0.790.830.810.950.950.780.950.830.900.870.87
Drink0.660.830.480.840.700.680.680.870.850.820.74
Answer0.791.000.680.890.650.860.940.960.800.780.84
Clap1.001.000.950.841.000.911.000.921.000.780.94
Tight0.970.940.951.000.950.860.950.921.000.950.95
Sit down0.720.890.900.900.760.860.791.000.800.910.85
Stand up1.000.831.000.900.900.900.840.880.950.960.92
Read watch0.590.890.900.840.750.820.680.750.850.730.78
Bow0.860.890.861.000.851.001.000.960.900.740.91
Average0.820.900.840.910.830.850.870.900.890.840.87