GitLab

Add support for DSVT

Add support for multi-modal NuScenes Detection

* Add files via upload

* Delete cbgs_voxel01_voxelnext_headkernel3.yaml

* Delete voxelnext_ioubranch.yaml

* VoxelNeXt

* Update .gitignore

* Update voxelnext_head.py

* Update spconv_utils.py

* add another slurm training script

* VoxelNeXt

* Fixbug: torch.cuda.amp rather than torch.amp.cuda

* use timm style to record log by logger
Co-authored-by: hova88 <yanhaixu@senior.auto>

* support amp
Co-authored-by: yukang <yukangchen@cse.cuhk.edu.hk>

* New CenterPoint parameters and scores

*  the offical support of ONCE dataset from ONCE team
Co-authored-by: JiagengMao <maojg@zju.edu.cn>
Co-authored-by: JiagengMao <maojiageng@gmail.com>

* grammatical errors

It will the keyerror 'gt_boxes' bug when evaluate on the test split, if we do not include this condition.

pillarnet: https://arxiv.org/abs/2205.07403