search

xinge008 / Cylinder3D

Rank 1st in the leaderboard of SemanticKITTI semantic segmentation (both single-scan and multi-scan) (Nov. 2020) (CVPR2021 Oral)
Apache License 2.0
859 stars 180 forks source link

A question #45

Closed jediofgever closed 3 years ago

jediofgever commented 3 years ago

Hello , I have read through the paper. The work looks fantastic. I was wondering whether this network is particularly suiting for my needs or its an overkill.

I have a pointcloud map that is automatically generated and labeled from simulation. I sematic segment the cloud to two class, ground and non-ground. The cloud doesn't have the features of LIDAR pointcloud such as sparsity in distant regions. My cloud is uniformly sampled almost equal density at any point, will this method(Cylinder3D ) still output state-of-the-art results for such a case ?

whats your thoughts ?

Thank you very much

xinge008 commented 3 years ago

In my opinion, you can use cylinder3D as the backbone network;

For your specific settings, ie, two-categories and closed relationship with height, you can design some modules to explore the geometric information (such as height) or design some specific loss functions. Just my two cents.