Regression-based human motion capture from voxel data

Conference item

A regression based method is proposed to recover human body pose from 3D voxel data. In order to do this we need to convert the voxel data into a feature vector. This is done using a Bayesian approach based on Mixture of Probabilistic PCA that transforms a collection of 3D shape context descriptors, extracted from the voxels, to a compact feature vector. For the regression, the newly-proposed Multi-Variate Relevance Vector Machine is explored to learn a single mapping from this feature vector to a low-dimensional representation of full body pose. We demonstrate the effectiveness and robustness of our method with experiments on both synthetic data and real sequences.

Authors: Sun, Y; Bray, M; Thayananthan, A; Yuan, B; Torr, PHS

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: British Machine Vision Association
• Host title: Proceedings of the British Machine Vision Conference 2006
• Pages: 277-286
• Publication date: 2006-01-01
• Event title: British Machine Vision Conference 2006
• Event location: Edinburgh
• Event website: https://bmva-archive.org.uk/bmvc/2006/
• Event start date: 2006-09-04
• Event end date: 2006-09-07
• Language: English