DUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scaling

1Universidad Torcuato Di Tella 2FCEyN Universidad de Buenos Aires 3CONICET
CVPR 2024

DUDF is able to leverage general-shape neural representation by learning a hyperbolic scaled unsigned distance field. The learning process is governed by solving a new and interesting Eikonal problem. This allows for detailed reconstructions, and extraction of important topological properties such as normal fields and shape curvatures; which were evasive in previous works.

Video

Abstract

In recent years, there has been a growing interest in training Neural Networks to approximate Unsigned Distance Fields (UDFs) for representing open surfaces in the context of 3D reconstruction. However, UDFs are non-differentiable at the zero level set which leads to significant errors in distances and gradients, generally resulting in fragmented and discontinuous surfaces. In this paper, we propose to learn a hyperbolic scaling of the unsigned distance field, which defines a new Eikonal problem with distinct boundary conditions. This allows our formulation to integrate seamlessly with state-of-the-art continuously differentiable implicit neural representation networks, largely applied in the literature to represent signed distance fields. Our approach not only addresses the challenge of open surface representation but also demonstrates significant improvement in reconstruction quality and training performance. Moreover, the unlocked field's differentiability allows the accurate computation of essential topological properties such as normal directions and curvatures, pervasive in downstream tasks such as rendering. Through extensive experiments, we validate our approach across various data sets and against competitive baselines. The results demonstrate enhanced accuracy and up to an order of magnitude increase in speed compared to previous methods.

Interpolate start reference image.

Baseline comparisons

We compare our method to state of the art neural representation approaches in three common open-surface datasets. Results show greater accuracy and improved training times.

Topological properties

The full differentiability of our method allows for mean and gaussian curvature computation.

Mean

Gaussian

Direct rendering and ilumination

Precise normal field and principal curvatures computation allows for realistic direct rendering techniques.

3D Demos

The triangle meshes extracted via gradient-based Marching Cubes have high detail and correct topology. Meshes have been simplified to run smoothly on web-browser.

BibTeX


      @InProceedings{Fainstein_2024_CVPR,
          author    = {Fainstein, Miguel and Siless, Viviana and Iarussi, Emmanuel},
          title     = {DUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scaling},
          booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
          month     = {June},
          year      = {2024},
          pages     = {4484-4493}
      }