Diagnosing and Fixing Manifold Overfitting in Deep Generative Models
This addresses a fundamental limitation in generative modeling for machine learning, providing a theoretical foundation and practical solution for improving density estimation.
The paper tackles the problem of manifold overfitting in deep generative models, where maximum-likelihood training fails to learn the distribution on low-dimensional data manifolds, and proposes a two-step procedure that avoids this issue and enables density estimation on manifolds learned by implicit models.
Likelihood-based, or explicit, deep generative models use neural networks to construct flexible high-dimensional densities. This formulation directly contradicts the manifold hypothesis, which states that observed data lies on a low-dimensional manifold embedded in high-dimensional ambient space. In this paper we investigate the pathologies of maximum-likelihood training in the presence of this dimensionality mismatch. We formally prove that degenerate optima are achieved wherein the manifold itself is learned but not the distribution on it, a phenomenon we call manifold overfitting. We propose a class of two-step procedures consisting of a dimensionality reduction step followed by maximum-likelihood density estimation, and prove that they recover the data-generating distribution in the nonparametric regime, thus avoiding manifold overfitting. We also show that these procedures enable density estimation on the manifolds learned by implicit models, such as generative adversarial networks, hence addressing a major shortcoming of these models. Several recently proposed methods are instances of our two-step procedures; we thus unify, extend, and theoretically justify a large class of models.