Skip to main content

Table 6 Results of our head pose estimation on the driver video sequence using 3 classes for pitch and yaw

From: Driver head pose estimation using efficient descriptor fusion

 

3 classes for pitch-SVM

3 classes for yaw-SVM

Database

Accuracy

Kappa

Time

Accuracy

Kappa

Time

Pointing’04

90.5

0.85

0.09

96.6

0.94

0.08

Our Sequence

97.5

0.96

0.03

98.2

0.98

0.02