Search papers, labs, and topics across Lattice.
This paper introduces a neural solver for computing Wasserstein geodesics and optimal transport (OT) maps between distributions. The method leverages the dynamical formulation of OT, recasting the constrained optimization as a minimax problem solved by deep neural networks approximating relevant functions. The approach recovers both the Wasserstein geodesic and the OT map, enabling direct sampling from the target distribution and estimation of the velocity field along particle trajectories.
Neural networks can now efficiently compute Wasserstein geodesics and optimal transport maps, offering a flexible approach applicable to various cost functions and datasets.
In recent years, the machine learning community has increasingly embraced the optimal transport (OT) framework for modeling distributional relationships. In this work, we introduce a sample-based neural solver for computing the Wasserstein geodesic between a source and target distribution, along with the associated velocity field. Building on the dynamical formulation of the optimal transport (OT) problem, we recast the constrained optimization as a minimax problem, using deep neural networks to approximate the relevant functions. This approach not only provides the Wasserstein geodesic but also recovers the OT map, enabling direct sampling from the target distribution. By estimating the OT map, we obtain velocity estimates along particle trajectories, which in turn allow us to learn the full velocity field. The framework is flexible and readily extends to general cost functions, including the commonly used quadratic cost. We demonstrate the effectiveness of our method through experiments on both synthetic and real datasets.
