Generative adversarial interpolative autoencoding: adversarial training on latent space interpolations encourage convex latent distributions.
Tim SainburgMarvin ThielkBrad TheilmanBenjamin MiglioriTimothy GentnerPublished in: CoRR (2018)
Keyphrases
- latent space
- generative model
- latent variables
- gaussian process latent variable models
- low dimensional
- latent variable models
- parameter space
- gaussian process
- gaussian processes
- lower dimensional
- high dimensional
- probabilistic model
- matrix factorization
- manifold learning
- gaussian mixture
- probabilistic latent semantic analysis
- feature space
- distance metric
- transfer learning
- random variables
- high dimensional spaces
- training set
- learning algorithm
- bayesian framework
- closed form
- principal component analysis
- training data