Uncovering Challenges of Solving the Continuous Gromov-Wasserstein Problem

2023-03-10

AI Summary

Existing methods for continuous Gromov–Wasserstein optimal transport (GWOT) implicitly rely on discretization and struggle to model parametric mappings between unknown continuous distributions, leading to fundamental deficiencies in theoretical modeling, numerical stability, and geometric fidelity. Method: We propose the first fully continuous GWOT framework—abandoning histogram approximations and grid-based constraints—and instead employ variational inference coupled with neural implicit mapping to learn differentiable, discretization-free parametric couplings. To ensure theoretical consistency with continuous OT, we integrate unbiased gradient estimation and adversarial distribution alignment. Contribution/Results: Experiments demonstrate that our method significantly outperforms state-of-the-art approaches in non-aligned, high-dimensional, and non-Euclidean manifold settings. It achieves superior mapping isometry and generalization on both synthetic and real-world data, establishing a differentiable and scalable theoretical and computational foundation for continuous GWOT.

Abstract

Recently, the Gromov-Wasserstein Optimal Transport (GWOT) problem has attracted the special attention of the ML community. In this problem, given two distributions supported on two (possibly different) spaces, one has to find the most isometric map between them. In the discrete variant of GWOT, the task is to learn an assignment between given discrete sets of points. In the more advanced continuous formulation, one aims at recovering a parametric mapping between unknown continuous distributions based on i.i.d. samples derived from them. The clear geometrical intuition behind the GWOT makes it a natural choice for several practical use cases, giving rise to a number of proposed solvers. Some of them claim to solve the continuous version of the problem. At the same time, GWOT is notoriously hard, both theoretically and numerically. Moreover, all existing continuous GWOT solvers still heavily rely on discrete techniques. Natural questions arise: to what extent do existing methods unravel the GWOT problem, what difficulties do they encounter, and under which conditions they are successful? Our benchmark paper is an attempt to answer these questions. We specifically focus on the continuous GWOT as the most interesting and debatable setup. We crash-test existing continuous GWOT approaches on different scenarios, carefully record and analyze the obtained results, and identify issues. Our findings experimentally testify that the scientific community is still missing a reliable continuous GWOT solver, which necessitates further research efforts. As the first step in this direction, we propose a new continuous GWOT method which does not rely on discrete techniques and partially solves some of the problems of the competitors.

Problem

• Identifying limitations in continuous GWOT solvers
• Assessing existing methods' effectiveness and challenges
• Proposing a new continuous GWOT solution method

Innovation

• Continuous Gromov-Wasserstein Optimal Transport
• Non-discrete technique application
• Benchmarking existing GWOT methods