VoxFormer: Sparse Voxel Transformer for Camera-Based 3D Semantic Scene Completion | IEEE Conference Publication | IEEE Xplore