Abstract
In this paper we approach the problem of head pose estimation by combining Multi-scale Gaussian Derivatives with Support Vector Machines. We evaluate the approach on the Pointing04 and CMU-PIE data sets and to estimate the pan and tilt of the head from facial images. We achieved a mean absolute error of 6.9 degrees for pan and 8.0 degrees for tilt on the Pointing04 data set. © 2013 Springer-Verlag.
Cite
CITATION STYLE
Jain, V., & Crowley, J. L. (2013). Head pose estimation using multi-scale Gaussian derivatives. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7944 LNCS, pp. 319–328). https://doi.org/10.1007/978-3-642-38886-6_31
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
