Complex Human Motion Estimation, FG 2008 (Conference Paper)

a presentation in Amsterdam.

T. Mukasa, A. Miyamoto, S. Nobuhara, A. Maki, and T. Matsuyama: Complex Human Motion Estimation Using Visibility, The 8th IEEE International Conference on Automatic Face and Gesture Recognition (FG2008), 2008. 9. (PDF)

概要：
３次元ビデオ中の複雑な人体運動の推定。
体節間の接触や自己遮蔽により、復元形状には信頼度の低い部位が存在する。
この点に着目し、我々は「可観測性」尺度を導入して部分形状の信頼度評価をすることで、頑健な形状マッチングを実現した。

Abstract:
A novel algorithm for estimating complex human motion from 3D video.
We base our algorithm on a model-based approach. This type of method usually works well against partly incomplete input data. However, it fails to estimate what we call “complex motion”: where some parts of the body touch each other for a long period.
In order to solve this problem, we introduce a “visibility” measure for each mesh vertex that represents how it is occluded or missed on the observed surface.
Using the “visibility” we selectively suppress the outliers caused by low observability while traditional surface matching algorithms try to find corresponding area for the entire surface
and cannot converge to the real posture by definition.