Xingguang (Kevin) Yan

TL;DR Hey guys, I have circumvented this error by just setting `aabb_scale=4` instead of `16` in the `transforms.json` file. --- More explanation: In my case of using V100 GPU, I...

It seems the result of `point_to_mesh_distance` is actually the squared unsigned distance. So you have to use the square root of the output as the real unsigned distance.

Thank you for your questions. 1. Yes, we apply scaling to the CD in the paper for better display. For Table 1 (completion), we scale CD by 10^3. For Table...

You can have a check at the Imnet2Dataset class at the file [ShapeFormer/shapeformer/data/imnet_datasets/imnet_datasets.py](https://github.com/QhelDIV/ShapeFormer/blob/3558e2a794c580795f465ca6ccd75a493850fda3/shapeformer/data/imnet_datasets/imnet_datasets.py#L350). There are some snippets to decompress the shape_vocab back to Ytg. Basically shape_vocab and vocab_idx is the...

For each demo shape, we store the complete shape in Xbd.npy and the partial point cloud in Xct.npy In general, you can provide only Xct.npy for testing, ignoreing Xbd.npy or...

Ah, you can use [mamba](https://mamba.readthedocs.io/en/latest/mamba-installation.html#mamba-install) instead, which is a community variant of conda which is much faster. The command is exactly conda command but faster. Or you can use the...

Yes, the pointnet module is retrained on our dataset.

It turns out this is due to the fact you are using a newer version of the xgutils library, where you can not import the utils by just `from xgutils...

Hi, thank you for your reply! The image comes from the gallery show cases in this repo's README.md, the url is: https://github.com/rougier/scientific-visualization-book/blob/master/images/scatter-3d.png Thank you again!

Thank you so much! I can directly render new point sets without any problem :)