Enhancement: radius

[tchaton]

Hey @rusty1s,

Check this repo: https://github.com/humanpose1/torch_radius_neighbors/tree/master
Nanoflann is faster / robust than scipy on cpu and remove the need to convert back from numpy to torch. Integrating the code might be interesting here :)
Also, do you have any idea on how to improve the radius search which is currently built on brute force search (it won't scale to large pointcloud)

Best,
Thomas Chaton

[rusty1s]

Thank you for the pointer. As an alternative GPU computation, we should benchmark a KeOps implementation against the current approach

[tchaton]

This paper seems pretty interesting for the task

semi_covex_hull_tree_icdm_accepted_final2.pdf

[tchaton]

Hey @rusty1s,

This could be used to make radius search faster on gpu.
http://on-demand.gputechconf.com/gtc/2014/presentations/S4117-fast-fixed-radius-nearest-neighbor-gpu.pdf

[github-actions]

This issue had no activity for 6 months. It will be closed in 2 weeks unless there is some new activity. Is this issue already resolved?