ML Anthology

A Semantic Occlusion Model for Human Pose Estimation from a Single Depth Image

Umer Rafi, Juergen Gall, Bastian Leibe
CVPRW 2015 pp. 67-74
doi:10.1109/CVPRW.2015.7301338 /cvprw/2015/rafi2015cvprw-semantic/

Abstract

Human pose estimation from depth data has made significant progress in recent years and commercial sensors estimate human poses in real-time. However, state-of-the-art methods fail in many situations when the humans are partially occluded by objects. In this work, we introduce a semantic occlusion model that is incorporated into a regression forest approach for human pose estimation from depth data. The approach exploits the context information of occluding objects like a table to predict the locations of occluded joints. In our experiments on synthetic and real data, we show that our occlusion model increases the joint estimation accuracy and outperforms the commercial Kinect 2 SDK for occluded joints.

CVPRW Semantic Scholar

Cite

Text

Rafi et al. "A Semantic Occlusion Model for Human Pose Estimation from a Single Depth Image." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2015. doi:10.1109/CVPRW.2015.7301338

Markdown

[Rafi et al. "A Semantic Occlusion Model for Human Pose Estimation from a Single Depth Image." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2015.](https://mlanthology.org/cvprw/2015/rafi2015cvprw-semantic/) doi:10.1109/CVPRW.2015.7301338

BibTeX

@inproceedings{rafi2015cvprw-semantic,
  title     = {{A Semantic Occlusion Model for Human Pose Estimation from a Single Depth Image}},
  author    = {Rafi, Umer and Gall, Juergen and Leibe, Bastian},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2015},
  pages     = {67-74},
  doi       = {10.1109/CVPRW.2015.7301338},
  url       = {https://mlanthology.org/cvprw/2015/rafi2015cvprw-semantic/}
}