Conversions across 3D Representations

Kaolin Library supports conversions between many 3D representations commonly used for deep learning with a plain PyTorch API.

From Mesh

• mesh to pointcloud:

  • kaolin.ops.mesh.sample_points

  • kaolin.ops.mesh.packed_sample_points

• mesh to voxels:

  • kaolin.ops.conversions.trianglemeshes_to_voxelgrids

• mesh to octree (Structured Point Clouds or SPC):

  • kaolin.ops.conversions.unbatched_mesh_to_spc

From Signed Distance Field (SDF)

• SDF to tetrahedral mesh:

  • kaolin.ops.conversions.marching_tetrahedra

• SDF to voxelgrid:

  • kaolin.ops.conversions.pointclouds_to_voxelgrids

From Point Cloud

• point cloud to voxels:

  • kaolin.ops.conversions.pointclouds_to_voxelgrids

• point cloud to octree (Structured Point Clouds or SPC):

  • kaolin.ops.conversions.unbatched_pointcloud_to_spc

From Voxels

• point cloud to mesh:

  • kaolin.ops.conversions.voxelgrids_to_cubic_meshes

  • kaolin.ops.conversions.voxelgrids_to_trianglemeshes

Hybrid Representations

• SDF and voxelgrid to tetrahedral mesh:

  • kaolin.ops.conversions.FlexiCubes (See also Volumetric Mesh Optimization)