Latent Variable Modeling for Generative Concept Representations and Deep Generative Models
arXiv:1812.11856
Abstract
Latent representations are the essence of deep generative models and determine their usefulness and power. For latent representations to be useful as generative concept representations, their latent space must support latent space interpolation, attribute vectors and concept vectors, among other things. We investigate and discuss latent variable modeling, including latent variable models, latent representations and latent spaces, particularly hierarchical latent representations and latent space vectors and geometry. Our focus is on that used in variational autoencoders and generative adversarial networks.
arXiv admin note: text overlap with arXiv:1706.00400 by other authors
References in corpus (4)
- Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling
- Multi-Level Variational Autoencoder: Learning Disentangled Representations from Grouped Observations
- Precise Recovery of Latent Vectors from Generative Adversarial Networks
- A Contemporary Overview of Probabilistic Latent Variable Models