Yi Wei
torch==1.10.1. cuda==11.1
Hi， 你试过在执行python setup.py install之前pip install urllib3==1.9吗
Hi, do you use the nuscenes data or your own data?
I doubt it maybe due to the pc range. -1 is too high for the z-axis lower bound of nuscenes data since the origin of LiDAR coordinate is on the...
It is based on BEVFormer's [config](https://github.com/fundamentalvision/BEVFormer/blob/2c889a36c71fa54534032b955ef0463526155951/projects/configs/bevformer/bevformer_base.py#L11).
Hi, you may try this command to generate pkl on your own: python tools/create_data.py nuscenes --root-path ./data/nuscenes --out-dir ./data/nuscenes --extra-tag nuscenes --version v1.0 --canbus ./data, which is same with [BEVFormer](https://github.com/fundamentalvision/BEVFormer/blob/master/docs/prepare_dataset.md)...
We do not try our model on their dataset. If you want to try, you should pay attention to the [data format](https://github.com/weiyithu/SurroundOcc/blob/main/docs/data.md), which is different with theirs.
Hi, thanks for your appreciation of our work. We do not conduct LiDAR segmentation experiments since we it cannot verify 3D reconstruction performance. If you want to do LiDAR segmentation,...
Hi, we do not test the case that sample_per_gpu > 1 and in our experiment we fix sample_pre_gpu as 1. I think the bug maybe caused by dataloader since the...
Hi, you mean the panoptic occupancy labels or using the panoptic images as inputs?