You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for releasing this good dataset!
When I checked the flows provided by the dataset, I found that the 'terrain' voxels also have flows. the average norm of the terrain flows is 1.67 in the training set and about 0.9 in the val set. I think the terrain should be a static class, so I can't understand why they have large flow values.
I downloaded the dataset from OpenDataLab.
The text was updated successfully, but these errors were encountered:
That's a error in the dataset. We have checked it and it only happened in the 'terrain' voxels. Please set them to 0 for a temporary solution. We will update the dataset later.
Thanks for releasing this good dataset!
When I checked the flows provided by the dataset, I found that the 'terrain' voxels also have flows. the average norm of the terrain flows is 1.67 in the training set and about 0.9 in the val set. I think the terrain should be a static class, so I can't understand why they have large flow values.
I downloaded the dataset from OpenDataLab.
The text was updated successfully, but these errors were encountered: