GLSR-VAE: Geodesic latent space regularization for variational autoencoder architectures.
Gaëtan HadjeresFrank NielsenFrançois PachetPublished in: SSCI (2017)
Keyphrases
- latent space
- latent variables
- low dimensional
- gaussian process
- generative model
- parameter space
- dimensionality reduction
- feature space
- gaussian processes
- manifold learning
- lower dimensional
- matrix factorization
- gaussian process latent variable models
- image segmentation
- distance metric
- transfer learning
- high dimensional
- probabilistic latent semantic analysis
- prior information
- high dimensional data
- gaussian mixture
- principal component analysis
- high dimensional spaces
- data sets
- image reconstruction
- distance transform
- pattern recognition