Quantitative comparison setting, and MVS-Synth dataset

[Byeongjun1022]

1. Can I know the setting for getting the same quantitative result with the paper?

2. Do you have a plan to provide the patchfusion network trained with MVS-Synth? or training code for training with MVS-Synth?
   As far as I see, 'splits/u4k/train.txt' only contains UnrealStereo4K dataset.

Thank you for releasing the code

[zhyever]

Could you please check 2d87adc with git check 2d87adc9? I think I provide all initial codes there, including the configs. Checkpoints are provided here: https://huggingface.co/zhyever/PatchFusion/tree/main. You can use the patchfusion_gta.pt and patchfusion_u4k.pt to reproduce the results in our paper.