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

Details request #171

Closed bhaba-ranjan closed 10 months ago

bhaba-ranjan commented 1 year ago

For NuScenes dataset, I see this in the readme.md file

  1. Pre-processing: we follow the data pre-processing in SECOND to prepare the nuscenes_info file, including nuscenes_infos_train.pkl, nuscenes_infos_val.pkl and nuscenes_infos_test.pkl.

Can someone please help me understand how this files are created? is there another repository with the code ref to generate this files?

YEOMJINSEOP commented 10 months ago

Did you figure out?

bhaba-ranjan commented 10 months ago

Negative @YEOMJINSEOP