Statistical Edge Detection And UDF Learning For Shape Representation

TL;DR

This paper introduces a statistical method to detect surface edges more accurately and a training approach that emphasizes these edges, resulting in improved neural UDFs for 3D surface representation.

Contribution

It presents a novel statistical edge detection technique and a focused sampling strategy to enhance the fidelity of neural UDFs in 3D shape modeling.

Findings

Enhanced surface edge detection accuracy

Improved neural UDF fidelity near edges

Better global shape representation

Abstract

In the field of computer vision, the numerical encoding of 3D surfaces is crucial. It is classical to represent surfaces with their Signed Distance Functions (SDFs) or Unsigned Distance Functions (UDFs). For tasks like representation learning, surface classification, or surface reconstruction, this function can be learned by a neural network, called Neural Distance Function. This network, and in particular its weights, may serve as a parametric and implicit representation for the surface. The network must represent the surface as accurately as possible. In this paper, we propose a method for learning UDFs that improves the fidelity of the obtained Neural UDF to the original 3D surface. The key idea of our method is to concentrate the learning effort of the Neural UDF on surface edges. More precisely, we show that sampling more training points around surface edges allows better local accuracy of the trained Neural UDF, and thus improves the global expressiveness of the Neural UDF in terms of Hausdorff distance. To detect surface edges, we propose a new statistical method based on the calculation of a $p$-value at each point on the surface. Our method is shown to detect surface edges more accurately than a commonly used local geometric descriptor.