Dimensionality Reduction of Collective Motion by Principal Manifolds.
Kelum GajamannageSachit ButailMaurizio PorfiriErik M. BolltPublished in: CoRR (2015)
Keyphrases
- dimensionality reduction
- manifold learning
- low dimensional
- high dimensional data
- high dimensional
- nonlinear dimensionality reduction
- image sequences
- principal component analysis
- motion estimation
- motion analysis
- feature space
- space time
- camera motion
- motion model
- low dimensional spaces
- human motion
- random projections
- high dimensionality
- laplacian eigenmaps
- feature extraction
- lie group
- pattern recognition and machine learning
- motion planning
- motion capture
- euclidean space
- lower dimensional
- manifold structure
- structure preserving
- neighborhood preserving
- data representation
- motion segmentation
- principal components
- linear discriminant analysis
- optical flow
- locally linear embedding
- feature selection
- embedding space
- euclidean distance
- data points
- metric learning
- least squares
- pattern recognition
- underlying manifold
- video sequences
- configuration space
- motion tracking
- motion field
- geometric structure