To train model with SceneFlow final pass dataset:
$ python run_train.py @inproceedings{kwon2023image,
title={Image-Coupled Volume Propagation for Stereo Matching},
author={Kwon, Oh-Hun and Zell, Eduard},
booktitle={2023 IEEE International Conference on Image Processing (ICIP)},
pages={2510--2514},
year={2023},
organization={IEEE}
}