Official implementation of the work "Computing high-dimensional optimal transport by flow neural networks"
conda env create -f environment.yml
conda activate FlowOT- Get the flow refinment given initial flows (stored in
checkpoints/)
python main_FlowOT.pyThe trajectory of the flow refinement is shown below:
- (Optional) Get the infinitesimal DRE given the flow refinment
python main_infinitesimal_DRE.py@inproceedings{xu2025computing,
title={Computing high-dimensional optimal transport by flow neural networks},
author={Chen Xu and Xiuyuan Cheng and Yao Xie},
booktitle={The 28th International Conference on Artificial Intelligence and Statistics},
year={2025},
url={https://openreview.net/forum?id=oEWYNesvRJ}
}