# Neural Convolutional Surfaces

Source: [Neural Convolutional Surfaces](<https://www.sota2.com/research/paper/68fa6c2c1e209131cea3285a>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-01T14:23:17.117Z

Authors: Luca Morreale, Noam Aigerman, Paul Guerrero, Vladimir G. Kim, Niloy J. Mitra

Published: 2022-04-05T15:40:11.000Z

## Abstract

This work is concerned with a representation of shapes that disentangles fine, local and possibly repeating geometry, from global, coarse structures. Achieving such disentanglement leads to two unrelated advantages: i) a significant compression in the number of parameters required to represent a given geometry; ii) the ability to manipulate either global geometry, or local details, without harming the other. At the core of our approach lies a novel pipeline and neural architecture, which are optimized to represent one specific atlas, representing one 3D surface. Our pipeline and architecture are designed so that disentanglement of global geometry from local details is accomplished through optimization, in a completely unsupervised manner. We show that this approach achieves better neural shape compression than the state of the art, as well as enabling manipulation and transfer of shape details. Project page at http://geometry.cs.ucl.ac.uk/projects/2022/cnnmaps/ .

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Neural Shape Representation on Armadillo](<https://www.sota2.com/research/sota/neural-shape-representation-on-armadillo.md>): Neural Shape Representation; Armadillo; Chamfer Distance: 0.54.

- [Neural Shape Representation on Bimba](<https://www.sota2.com/research/sota/neural-shape-representation-on-bimba.md>): Neural Shape Representation; Bimba; Chamfer Distance: 1.04.

- [Neural Shape Representation on Dino](<https://www.sota2.com/research/sota/neural-shape-representation-on-dino.md>): Neural Shape Representation; Dino; Chamfer Distance: 1.48.

- [Neural Shape Representation on Dragon](<https://www.sota2.com/research/sota/neural-shape-representation-on-dragon.md>): Neural Shape Representation; Dragon; Chamfer Distance: 0.57.

- [Neural Shape Representation on Seahorse](<https://www.sota2.com/research/sota/neural-shape-representation-on-seahorse.md>): Neural Shape Representation; Seahorse; Chamfer Distance: 0.44.

- [Neural Shape Representation on Gargoyl](<https://www.sota2.com/research/sota/neural-shape-representation-on-gargoyl.md>): Neural Shape Representation; Gargoyl; Chamfer Distance: 2.29.

- [Neural Shape Representation on Elephant](<https://www.sota2.com/research/sota/neural-shape-representation-on-elephant.md>): Neural Shape Representation; Elephant; Chamfer Distance: 2.49.

- [Neural Shape Representation on Grog](<https://www.sota2.com/research/sota/neural-shape-representation-on-grog.md>): Neural Shape Representation; Grog; Chamfer Distance: 1.28.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2204.02289>)

[PDF](<https://arxiv.org/pdf/2204.02289.pdf>)

[Implementation](<http://geometry.cs.ucl.ac.uk/projects/2022/cnnmaps/>)