Fast Online Upper Body Pose Estimation from Video

Ming-Ching Chang, Honggang Qi, Xin Wang, Hong Cheng and Siwei Lyu

Abstract

Estimation of human body poses from video is an important problem in computer vi- sion with many applications. Most existing methods for video pose estimation are offline in nature, where all frames in the video are used in the process to estimate the body pose in each frame. In this work, we describe a fast online video upper body pose estima- tion method (CDBN-MODEC) that is based on a conditional dynamic Bayesian network model, which predicts upper body pose in a frame without using information from fu- ture frames. Our method combines fast single image based pose estimation methods with the temporal correlation of poses between frames. We collect a new high frame rate upper body pose dataset that better reflects practical scenarios calling for fast online video pose estimation. When evaluated on this dataset and the VideoPose2 benchmark dataset, CDBN-MODEC achieves improvements in both performance and running effi- ciency over several state-of-art online video pose estimation methods.

Session

Poster 2

Files

Extended Abstract (PDF, 839K)
Paper (PDF, 3M)

DOI

10.5244/C.29.104
https://dx.doi.org/10.5244/C.29.104

Citation

Bibtex

@inproceedings{BMVC2015_104,
	title={Fast Online Upper Body Pose Estimation from Video},
	author={Ming-Ching Chang and Honggang Qi and Xin Wang and Hong Cheng and Siwei Lyu},
	year={2015},
	month={September},
	pages={104.1-104.12},
	articleno={104},
	numpages={12},
	booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
	publisher={BMVA Press},
	editor={Xianghua Xie, Mark W. Jones, and Gary K. L. Tam},
	doi={10.5244/C.29.104},
	isbn={1-901725-53-7},
	url={https://dx.doi.org/10.5244/C.29.104}
}