A second-order PDE technique to construct distance functions with more accurate derivatives.
Siddharth ManayAnthony J. YezziPublished in: ICIP (1) (2003)
Keyphrases
- distance function
- higher order
- nearest neighbor
- similarity search
- knn
- high order
- euclidean distance
- distance measure
- metric space
- differential operators
- partial differential equations
- similarity function
- finite difference
- quadratic form
- distance metric
- neural network
- directional derivatives
- k nearest neighbor
- anisotropic diffusion
- data points
- fourth order
- multiscale
- machine learning
- image denoising
- numerical solution
- distance computation
- face recognition