search

engineerJPark / LiDARWeather

[ECCV 2024 Oral] Official code of "Rethinking Data Augmentation for Robust LiDAR Semantic Segmentation in Adverse Weather".
Apache License 2.0
38 stars 2 forks source link

Where is the .minkunet_backbone_batchnormmixer? I can't find it. #5

Closed fdy61 closed 2 months ago

engineerJPark commented 2 months ago

The minkunet_backbone_batchnormmixer was abandoned during past experiments and is no longer used. You can go ahead and delete that part from the init files, or you can wait for me to commit the updated code. I'm currently refactoring and debugging everything before the conference day.

fdy61 commented 2 months ago

Thank you. And Another question. When I run the command of create_data.py, there is an error of ImportError: cannot import name 'RandomJitterPoints' from 'mmdet3d.datasets.transforms.transforms_3d'. I have checked it and find that RandomJitterPoints module is not in it indeed.

engineerJPark commented 2 months ago

As mentioned in other issue, I mistakenly deleted that part while refactoring the code. You can find it in a previous commit, or if you prefer, you can wait for me to commit the updated version. I'm currently refactoring and debugging everything before the conference day.

engineerJPark commented 2 months ago

As I mentioned in here right before, I have fixed the issues and checked that the code runs correctly in the latest commit. Please check for this.

fdy61 commented 2 months ago

Thank you very much!