Learning elementary structures for 3D shape generation and matching
This work addresses the problem of generating and matching 3D shapes for computer vision and graphics applications, offering incremental improvements over existing methods.
The paper tackles 3D shape generation and matching by representing shapes as deformations and combinations of learnable elementary 3D structures, achieving a 16% improvement in shape reconstruction over surface deformation approaches and a 6% improvement in state-of-the-art for dense correspondence estimation on the FAUST inter challenge.
We propose to represent shapes as the deformation and combination of learnable elementary 3D structures, which are primitives resulting from training over a collection of shape. We demonstrate that the learned elementary 3D structures lead to clear improvements in 3D shape generation and matching. More precisely, we present two complementary approaches for learning elementary structures: (i) patch deformation learning and (ii) point translation learning. Both approaches can be extended to abstract structures of higher dimensions for improved results. We evaluate our method on two tasks: reconstructing ShapeNet objects and estimating dense correspondences between human scans (FAUST inter challenge). We show 16% improvement over surface deformation approaches for shape reconstruction and outperform FAUST inter challenge state of the art by 6%.