# Unsupervised cycle-consistent deformation for shape matching

## TL;DR

This paper introduces a self-supervised deep learning method for shape matching that uses cycle-consistency as a supervisory signal, enabling effective deformation transfer without relying on templates or extensive annotations.

## Contribution

It presents a novel cycle-consistency based self-supervised approach for shape deformation and matching, outperforming existing methods especially in few-shot scenarios.

## Key findings

- Competitive with state-of-the-art when ample data is available.
- Significantly better in few-shot segmentation tasks.
- Does not require templates or point-correspondence supervision.

## Abstract

We propose a self-supervised approach to deep surface deformation. Given a pair of shapes, our algorithm directly predicts a parametric transformation from one shape to the other respecting correspondences. Our insight is to use cycle-consistency to define a notion of good correspondences in groups of objects and use it as a supervisory signal to train our network. Our method does not rely on a template, assume near isometric deformations or rely on point-correspondence supervision. We demonstrate the efficacy of our approach by using it to transfer segmentation across shapes. We show, on Shapenet, that our approach is competitive with comparable state-of-the-art methods when annotated training data is readily available, but outperforms them by a large margin in the few-shot segmentation scenario.