colmap_to_mega_nerf.py what exactly is the scale parameter in this scale

[ghaodong]

Thank you for such an excellent work, When I use my own data set to convert from colmap format to mega-nerf format, I colmap_to_mega_nerf.py what exactly is the scale parameter in this scale, I don't understand here, how much should I set appropriately?

[hturki]

the idea to apply a scale parameter which will divide all of your camera positions so that they fall within the [-1, 1] range

[ghaodong]

the idea to apply a scale parameter which will divide all of your camera positions so that they fall within the [-1, 1] range

Okay, thanks for the reply, in addition, in the create_cluster_masks.py, grid_dim this parameter, I see that it is int type, so I set it to 8, but the error index is out of range, and I see that it is a 2-dimensional in this, so how to set this parameter。

[hturki]

grid_dim is used to specify the dimension of the grid that's used to split up the scene into submodules. For example, if you want to split up the world into a 4x2 grid (and train a separate nerf submodule for each), use grid_dim 4 2