Giters

Pointcept / Pointcept

Pointcept: a codebase for point cloud perception research. Latest works: PTv3 (CVPR'24 Oral), PPT (CVPR'24), OA-CNNs (CVPR'24), MSC (CVPR'23)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Question about grid_sample and multi-frame

pycoco opened this issue · comments

commented

  1. why to have two grid_sample in test configuration. And the mode of first one is "train"?
    https://github.com/Pointcept/Pointcept/blob/main/configs/waymo/semseg-pt-v3m1-0-base.py#L172
    image

  2. if i want to use multi-frame for ptv3, how should i to modify the code? thanks!

commented

why to have two grid_sample in test configuration. And the mode of first one is "train"?

A faster testing strategy which mixes interpolation (original -> 0.025) and our precise testing (0.025 -> 0.05)

if I want to use multi-frame for ptv3, how should I modify the code?

  1. For the Waymo dataset, I will have an update to support multi-frames later.
  2. For semantic-kitti, I write a version here https://github.com/CVMI-Lab/MarS3D/blob/main/pointseg/datasets/semantic_kitti.py
commented

thanks very much! and is there a ptv3 config for lidar detection? I add a head after encode layers, but map is always to zero. @Gofinge