Deformable Surface Reconstruction via Riemannian Metric Preservation

TL;DR

This paper introduces a neural network-based method for reconstructing deformable surfaces from image sequences by preserving Riemannian metrics, achieving state-of-the-art results without offline training.

Contribution

It proposes a novel approach leveraging Riemannian metric preservation in neural networks for surface reconstruction from monocular images, bypassing the need for offline training.

Findings

Achieves state-of-the-art performance on benchmark datasets.

Does not require offline training, enabling real-time applications.

Effectively infers continuous deformable surfaces from image sequences.

Abstract

Estimating the pose of an object from a monocular image is an inverse problem fundamental in computer vision. The ill-posed nature of this problem requires incorporating deformation priors to solve it. In practice, many materials do not perceptibly shrink or extend when manipulated, constituting a powerful and well-known prior. Mathematically, this translates to the preservation of the Riemannian metric. Neural networks offer the perfect playground to solve the surface reconstruction problem as they can approximate surfaces with arbitrary precision and allow the computation of differential geometry quantities. This paper presents an approach to inferring continuous deformable surfaces from a sequence of images, which is benchmarked against several techniques and obtains state-of-the-art performance without the need for offline training.