Skip to Main content Skip to Navigation
Reports

Learning Methods for Recovering 3D Human Pose from Monocular Images

Ankur Agarwal 1 Bill Triggs 1
1 LEAR - Learning and recognition in vision
GRAVIR - IMAG - Laboratoire d'informatique GRAphique, VIsion et Robotique de Grenoble, Inria Grenoble - Rhône-Alpes, CNRS - Centre National de la Recherche Scientifique : FR71
Abstract : We describe a learning based method for recovering 3D human body pose from single images and monocular image sequences. Our approach requires neither an explicit body model nor prior labelling of body parts in the image. Instead, it recovers pose by direct nonlinear regression against shape descriptor vectors extracted automatically from image silhouettes. For robustness against local silhouette segmentation errors, silhouette shape is encoded by histogram-of-shape-contexts descriptors. We evaluate several different regression methods: ridge regression, Relevance Vector Machine (RVM) regression and Support Vector Machine (SVM) regression over both linear and kernel bases. The RVMs provide much sparser regressors without compromising performance, and kernel bases give a small but worthwhile improvement in performance. Loss of depth and limb labelling information often makes the recovery of 3D pose from single silhouettes ambiguous. We propose two solutions to this: the first embeds the method in a tracking framework, using dynamics from the previous state estimate to disambiguate the pose; the second uses a mixture of regressors framework to return multiple solutions for each silhouette. We show that the resulting system tracks long sequences stably, and is also capable of accurately reconstructing 3D human pose from single images, giving multiple possible solutions in ambiguous cases. For realism and good generalization over a wide range of viewpoints, we train the regressors on images resynthesized from real human motion capture data. The method is demonstrated on a 54-parameter full body pose model, both quantitatively on independent but similar test data, and qualitatively on real image sequences. Mean angular errors of 4-5 degrees are obtained - a factor of 3 better than the current state of the art for the much simpler upper body problem.
Document type :
Reports
Complete list of metadata

Cited literature [33 references]  Display  Hide  Download


https://hal.inria.fr/inria-00070668
Contributor : Rapport de Recherche Inria <>
Submitted on : Friday, May 19, 2006 - 9:09:55 PM
Last modification on : Monday, December 28, 2020 - 3:44:02 PM
Long-term archiving on: : Monday, September 17, 2012 - 4:01:21 PM

Identifiers

  • HAL Id : inria-00070668, version 1

Collections

Citation

Ankur Agarwal, Bill Triggs. Learning Methods for Recovering 3D Human Pose from Monocular Images. [Research Report] RR-5333, INRIA. 2004, pp.25. ⟨inria-00070668⟩

Share

Metrics

Record views

432

Files downloads

640