Using Segmented 3D Point Clouds for Accurate Likelihood Approximation in Human Pose Tracking.
Nicolas H. LehmentMoritz KaiserGerhard RigollPublished in: Int. J. Comput. Vis. (2013)
Keyphrases
- human pose tracking
- pose estimation
- high quality
- computationally efficient
- maximum likelihood
- information systems
- high accuracy
- real time
- high precision
- structured light
- data sets
- three dimensional
- multiscale
- social networks
- machine learning
- depth map
- range images
- point cloud
- error bounds
- approximation methods
- likelihood model