SVGD as a kernelized Wasserstein gradient flow of the chi-squared divergence.
Sinho ChewiThibaut Le GouicChen LuTyler MaunuPhilippe RigolletPublished in: CoRR (2020)
Keyphrases
- chi squared
- gradient flow
- active contours
- information gain
- energy functional
- partial differential equations
- active contour model
- level set
- variational formulation
- curve evolution
- image dependent
- level set method
- energy minimization
- kernel function
- feature selection
- mumford shah functional
- image segmentation
- object boundaries
- shape prior
- segmentation method
- energy function
- higher order
- bayesian networks
- feature extraction