Extremal Domain Translation with Neural Optimal Transport

2023/01/30 by Gazdieva, Milena, Korotin, Alexander, Selikhanovych, Daniil +1 · 1 citation
#FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2301.12874

Abstract

In many unpaired image domain translation problems, e.g., style transfer or super-resolution, it is important to keep the translated image similar to its respective input image. We propose the extremal transport (ET) which is a mathematical formalization of the theoretically best possible unpaired translation between a pair of domains w.r.t. the given similarity function. Inspired by the recent advances in neural optimal transport (OT), we propose a scalable algorithm to approximate ET maps as a limit of partial OT maps. We test our algorithm on toy examples and on the unpaired image-to-image translation task. The code is publicly available at https://github.com/milenagazdieva/ExtremalNeuralOptimalTransport

Cited by

• Learning of Population Dynamics: Inverse Optimization Meets JKO Scheme

Related

• Improving Neural Optimal Transport via Displacement Interpolation
• Overcoming Spurious Solutions in Semi-Dual Neural Optimal Transport: A Smoothing Approach for Learning the Optimal Transport Plan
• Large-Scale Optimal Transport via Adversarial Training with Cycle-Consistency
• Unbalancedness in Neural Monge Maps Improves Unpaired Domain Translation
• OT-Net: A Reusable Neural Optimal Transport Solver