This is the official PyTorch implementation of object detection for SphereFormer (CVPR 2023).
Spherical Transformer for LiDAR-based 3D Recognition [Paper]
Xin Lai, Yukang Chen, Fanbin Lu, Jianhui Liu, Jiaya Jia
Clone the repo via
git clone https://github.com/dvlab-research/SphereFormer.git --recursive && cd SphereFormer/detection/
This implementataion is built on OpenPCDet (https://github.com/open-mmlab/OpenPCDet). Please strictly follow its official guidance for installation and data preparation.
In addition to installing OpenPCDet, follow the following command to install Sparse Transformer (SpTr).
cd tools/third_party/SparseTransformer && python setup.py install
bash scripts/dist_train.sh 4 --cfg_file cfgs/nuscenes_models/cbgs_voxel0075_res3d_centerpoint_sphereformer.yaml