Supervised deep learning of elastic SRV distances on the shape space of curves

TL;DR

This paper presents a supervised deep learning method to efficiently compute SRV distances between curves, bypassing complex optimization procedures, with demonstrated improvements in speed and accuracy across various experiments.

Contribution

It introduces a novel deep learning framework for directly estimating SRV distances, enhancing computational efficiency and accuracy in shape analysis tasks.

Findings

Deep learning approach significantly reduces computation time.

Method achieves higher accuracy than traditional optimization-based methods.

Numerical experiments validate effectiveness across applications.

Abstract

Motivated by applications from computer vision to bioinformatics, the field of shape analysis deals with problems where one wants to analyze geometric objects, such as curves, while ignoring actions that preserve their shape, such as translations, rotations, or reparametrizations. Mathematical tools have been developed to define notions of distances, averages, and optimal deformations for geometric objects. One such framework, which has proven to be successful in many applications, is based on the square root velocity (SRV) transform, which allows one to define a computable distance between spatial curves regardless of how they are parametrized. This paper introduces a supervised deep learning framework for the direct computation of SRV distances between curves, which usually requires an optimization over the group of reparametrizations that act on the curves. The benefits of our approach in terms of computational speed and accuracy are illustrated via several numerical experiments.