A single-network fixed-point formulation for neural optimal transport eliminates adversarial min-max optimization and implicit differentiation while enforcing dual feasibility exactly.
(2021), arXiv: 2102.02992
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
math.OC 3verdicts
UNVERDICTED 3representative citing papers
A new constrained gradient flow on the space of transport maps converges to the OT map and enables more stable and accurate training of convexity-constrained neural networks for learning Monge maps.
SOCP reformulation of staggered-grid dynamic optimal transport eliminates interpolation steps and enables efficient proximal augmented Lagrangian solving with demonstrated speed and robustness gains.
citing papers explorer
-
Fixed-Point Neural Optimal Transport without Implicit Differentiation
A single-network fixed-point formulation for neural optimal transport eliminates adversarial min-max optimization and implicit differentiation while enforcing dual feasibility exactly.
-
Learning Monge maps with constrained drifting models
A new constrained gradient flow on the space of transport maps converges to the OT map and enables more stable and accurate training of convexity-constrained neural networks for learning Monge maps.
-
An efficient second-order cone programming approach for dynamic optimal transport on staggered grid discretization
SOCP reformulation of staggered-grid dynamic optimal transport eliminates interpolation steps and enables efficient proximal augmented Lagrangian solving with demonstrated speed and robustness gains.