A Wasserstein-Type Distance for Gaussian Mixtures on Vector Bundles with Applications to Shape Analysis.
Michael WilsonTom NeedhamChiwoo ParkSuparteek KunduAnuj SrivastavaPublished in: SIAM J. Imaging Sci. (2024)
Keyphrases
- shape analysis
- gaussian mixture
- shape representation
- kl divergence
- shape descriptors
- image analysis and pattern recognition
- gaussian mixture model
- medical imaging
- covariance matrix
- expectation maximization
- closed form
- probability density function
- em algorithm
- covariance matrices
- geodesic distance
- pointwise
- image processing
- markov random field
- three dimensional
- machine learning
- vector space
- euclidean distance
- mixture model
- distance measure
- probability distribution
- object recognition
- shape space