Geodesic Convolutional Neural Networks on Riemannian Manifolds.
Jonathan MasciDavide BoscainiMichael M. BronsteinPierre VandergheynstPublished in: ICCV Workshops (2015)
Keyphrases
- riemannian manifolds
- convolutional neural networks
- geodesic distance
- euclidean space
- riemannian metric
- shortest path
- manifold learning
- euclidean distance
- distance transform
- geometric structure
- distance metric
- shape space
- vector space
- metric space
- multi class classification
- parameter space
- mean shift
- shape analysis
- tangent space
- feature space
- feature selection
- covariance matrix
- gaussian mixture model
- low dimensional
- dynamic programming