Kernel PLS Regression for Robust Monocular Pose Estimation

Abstract

We evaluate the robustness of five regression techniques for monocular 3D pose estimation. While most of the discriminative pose estimation methods focus on overcoming the fundamental problem of insufficient training data, we are interested in characterizing performance improvement for increasingly large training sets. Commercially available rendering software allows us to efficiently generate large numbers of realistic images of poses from diverse actions. Inspired by recent work in human detection, we apply PLS and kPLS regression to pose estimation. We observe that kPLS regression incrementally approximates GP regression using the strongest nonlinear correlations between image features and pose. This provides robustness, and our experiments show kPLS regression is more robust than two GP-based state-of-the-art methods for pose estimation.

Cite

Text

Dondera and Davis. "Kernel PLS Regression for Robust Monocular Pose Estimation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2011. doi:10.1109/CVPRW.2011.5981750

Markdown

[Dondera and Davis. "Kernel PLS Regression for Robust Monocular Pose Estimation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2011.](https://mlanthology.org/cvprw/2011/dondera2011cvprw-kernel/) doi:10.1109/CVPRW.2011.5981750

BibTeX

@inproceedings{dondera2011cvprw-kernel,
  title     = {{Kernel PLS Regression for Robust Monocular Pose Estimation}},
  author    = {Dondera, Radu and Davis, Larry},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2011},
  pages     = {24-30},
  doi       = {10.1109/CVPRW.2011.5981750},
  url       = {https://mlanthology.org/cvprw/2011/dondera2011cvprw-kernel/}
}