facebookresearch / facebookresearch/pytorch3d
furthest_point_sample is slow when applied to large point clouds.
Open
- Dominant language
- Python
- Stars
- 10k
- Forks
- 1.5k
- PR merge metrics
- No merged PRs in 30d
Description
_furthest_point_sample_ works fine for small point clouds, but on larger ones (size > 300k) it becomes very slow—around 3× slower than the original PointNet version.
Contributor guide
Assessment
This issue has not been assessed yet.