Semantic-aware 3D-voxel CenterNet for point cloud object detection All code will be released upon publication of the paper. More detailed results of our SA-voxel-centernet are on the sa-voxel-centernet src file can get offset as this