Potential Flow Generator with $L_2$ Optimal Transport Regularity for Generative Models

Abstract

We propose a potential flow generator with $L_2$ optimal transport regularity, which can be easily integrated into a wide range of generative models including different versions of GANs and flow-based models. We show the correctness and robustness of the potential flow generator in several 2D problems, and illustrate the concept of "proximity" due to the $L_2$ optimal transport regularity. Subsequently, we demonstrate the effectiveness of the potential flow generator in image translation tasks with unpaired training data from the MNIST dataset and the CelebA dataset.

References in corpus (6)

• Spectral Normalization for Generative Adversarial Networks
• Variational Inference with Normalizing Flows
• Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations
• Learning to Discover Cross-Domain Relations with Generative Adversarial Networks
• Adversarial Computation of Optimal Transport Maps
• Scalable Unbalanced Optimal Transport using Generative Adversarial Networks

Cited by in corpus (2)

• A Machine Learning Framework for Solving High-Dimensional Mean Field Game and Mean Field Control Problems
• A Neural Network Approach for High-Dimensional Optimal Control Applied to Multi-Agent Path Finding