Search papers, labs, and topics across Lattice.
This paper introduces P-CORE, a self-supervised method that enhances point-based neural representations for non-rigid shape editing by ensuring surface consistency during large deformations. By generating random deformations and enforcing that surface predictions remain consistent before and after deformation, the method effectively addresses issues of holes and discontinuities that plague traditional point-based approaches. Experimental results show that P-CORE outperforms existing methods in zero-shot editing and significantly reduces artifacts, demonstrating its robustness in both synthetic and real-world scenarios.
Point-based neural editing can now adapt to large deformations without ground truth images, achieving superior consistency and fewer artifacts.
Advances in neural rendering have enabled high-fidelity multi-view reconstruction of 3D scenes. However, free-form non-rigid shape editing remains a significant challenge. Point-based neural representations are highly desirable for multi-view reconstruction because they lack fixed connectivity, which does not constrain the learned surface topology to that of the initialization. Yet this same property causes point-based representations to struggle with holes and surface discontinuities under large deformations. To address this, we propose a novel self-supervised method to enable point-based representations to adapt to large deformations without requiring ground truth multi-view images of deformed geometry. The key idea is to generate random deformations and to ensure consistency in the predicted surface before and after deformation. In particular, the surface prediction from the deformed point cloud should be the same as the deformation applied to the surface prediction from the original point cloud. We incorporate our approach into attention-based point representations, which differ from splatting-based point representations in their use of a learned interpolation kernel between points as opposed to a Gaussian kernel around each point. This learned interpolation kernel can learn to adapt to large deformations, without requiring addition or removal of points. We show that our framework significantly enhances its robustness to large deformations. Experiments on synthetic geometry editing benchmarks (Neural Editor, Objaverse) demonstrate that our approach outperforms existing point-based methods in zero-shot editing and significantly reduces artifacts. Furthermore, qualitative results on the DTU and Mip-NeRF 360 datasets demonstrate our method's effectiveness on real-world scenes.
