Different visualization result in demo_depth_map.ipynb about ditto

[Steve-Tod]

Hi, I just found out we used a different model weight to get the real-world results. It's trained for a longer time. It does not give a very good quantitative result, but it seems to generalize better to real-world scenarios without the need to tunning the marching cube parameters. The model is here. I'll also update it in the README.

[123zhen123]

Hi, I just found out we used a different model weight to get the real-world results. It's trained for a longer time. It does not give a very good quantitative result, but it seems to generalize better to real-world scenarios without the need to tunning the marching cube parameters. The model is here. I'll also update it in the README.

Thank you very much for your answer. I have the same problem. But this link seems to have been defunct, could you please send it again?

Meanwhile, I used "real_laptop.zip" data in "demo_depth_map.ipynb", but in IN [7] , I got incomplete point cloud. This looks different from the results provided. Is there some discrepancy in the data?