Closed Yacovitch closed 4 months ago
Hi, yes we do downsampling from the ground truth. The main reason to downsample to 180000 points is to fit it into GPU memory, otherwise the training and the denoising inference would just be intractable. That is why we use the refinement network to upsample the diffusion prediction afterward.
But also, usually the point cloud from SemanticKITTI do have around 180000 points as well.
Hi again,
In my understanding, scans in SemanticKITTI have an average 480,000 points, but in your code and configurations, you used 180000 points for aggregated input
p_full
and 18000 points forp_part
. Suppose I understand correctly, aggregated point cloud (from map_clean) will have a lot more points, and even considering filtering out moving objects and filtering points based on maximum range. In that case, the current configuration requires extreme downsampling. Is there any reason you picked 180000 as yournum_points
and 1/10 as your downsampling rate forp_part
?