Generative Modeling by Minimizing the Wasserstein-2 Loss

📅 2024-06-19
🏛️ arXiv.org
📈 Citations: 4
Influential: 0
📄 PDF
🤖 AI Summary
This work addresses the challenge of minimizing the Wasserstein-2 (W₂) distance in unsupervised generative modeling. We propose the first explicit, distribution-dependent ordinary differential equation (ODE) characterizing the W₂ gradient flow, and theoretically prove that its time-marginal distributions converge rigorously to the target data distribution. Methodologically, we design a persistent Euler discretization algorithm that couples with the gradient flow structure, circumventing the instability inherent in conventional adversarial training. Our key contributions are: (1) the first explicit ODE formulation of the W₂ gradient flow; and (2) a novel persistence training mechanism that enhances discretization fidelity and convergence robustness. Experiments demonstrate that our approach significantly outperforms WGAN on both high- and low-dimensional benchmarks; moreover, increasing persistence strength further improves generation quality and training stability.

Technology Category

Application Category

📝 Abstract
This paper approaches the unsupervised learning problem by minimizing the second-order Wasserstein loss (the $W_2$ loss) through a distribution-dependent ordinary differential equation (ODE), whose dynamics involves the Kantorovich potential associated with the true data distribution and a current estimate of it. A main result shows that the time-marginal laws of the ODE form a gradient flow for the $W_2$ loss, which converges exponentially to the true data distribution. An Euler scheme for the ODE is proposed and it is shown to recover the gradient flow for the $W_2$ loss in the limit. An algorithm is designed by following the scheme and applying persistent training, which naturally fits our gradient-flow approach. In both low- and high-dimensional experiments, our algorithm outperforms Wasserstein generative adversarial networks by increasing the level of persistent training appropriately.
Problem

Research questions and friction points this paper is trying to address.

Minimizes Wasserstein-2 loss for unsupervised learning
Uses ODE gradient flow to converge to data distribution
Outperforms WGANs with persistent training in experiments
Innovation

Methods, ideas, or system contributions that make the work stand out.

Minimizing Wasserstein-2 loss via distribution-dependent ODE
Using Euler scheme to recover gradient flow for convergence
Applying persistent training to enhance algorithm performance
🔎 Similar Papers
No similar papers found.