One problem in appearance-based pose estimation is the need of many training examples, i.e. images of the object in a large number of known poses. Some invariance can be obtained by considering translations, rotations and scale changes in the image plane, but the remaining degrees of freedom are often handled simply by sampling the pose space densely enough. This work presents a method for accurate interpolation between training views using local linear models. As a view representation local soft orientation histograms are used. The derivative of this representation with respect to the image plane transformations is computed, and a Gauss-Newton optimization is used to optimize all pose parameters simultaneously, resulting in an accurate estimate.
This paper is a compacted version of the SCIA paper with the same name. Please refer to this paper for a full-text version .
@inproceedings{jf07a,
Author = {Jonsson, E. and Felsberg, M.},
Booktitle = {SSBA},
Date-Added = {2007-08-07 11:14:41 +0200},
Date-Modified = {2007-08-07 11:14:41 +0200},
Title = {Accurate Interpolation in Appearance-Based Pose Estimation},
Year = {2007}}