Daniel-xsy / RoboBEV

RoboBEV: Towards Robust Bird's Eye View Perception under Common Corruption and Domain Shift
https://daniel-xsy.github.io/robobev/
326 stars 32 forks source link

Are there plans to support the occ benckmark used for the cvpr2023 competition? #10

Open SepJourney opened 8 months ago

SepJourney commented 8 months ago

For the occupancy task, the data and evaluation used for the cvpr2023 competition (https://github.com/CVPR2023-3D-Occupancy-Prediction/CVPR2023-3D-Occupancy-Prediction) are more popular, does this repository have plans to use it for the robust occupancy task?

Daniel-xsy commented 8 months ago

Thanks for your attention. we are sorry currently we don't have plans to update this. If we plan to update it in the future, we will let you know.

SepJourney commented 8 months ago

Thanks for your attention. we are sorry currently we don't have plans to update this. If we plan to update it in the future, we will let you know.

Thanks for your reply! Does this mean that you will follow the same evaluation protocol of TPVFormer listed in the README doc for the robust occupancy prediction challenge (ICRA2024)? (pseudo voxel label from lidar points)

Daniel-xsy commented 8 months ago

We will follow the same evaluation protocol of SurroundOcc for the RoboDrive challenge. Please refer to this link for more details. We will use the generated occupancy labels with resolution 200x200x16.