Grounding Continuous Representations in Geometry: Equivariant Neural Fields

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wessels, David R, Knigge, David M, Papa, Samuele, Valperga, Riccardo, Vadgama, Sharvaree, Gavves, Efstratios, Bekkers, Erik J
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929701540331520
author Wessels, David R
Knigge, David M
Papa, Samuele
Valperga, Riccardo
Vadgama, Sharvaree
Gavves, Efstratios
Bekkers, Erik J
author_facet Wessels, David R
Knigge, David M
Papa, Samuele
Valperga, Riccardo
Vadgama, Sharvaree
Gavves, Efstratios
Bekkers, Erik J
contents Conditional Neural Fields (CNFs) are increasingly being leveraged as continuous signal representations, by associating each data-sample with a latent variable that conditions a shared backbone Neural Field (NeF) to reconstruct the sample. However, existing CNF architectures face limitations when using this latent downstream in tasks requiring fine-grained geometric reasoning, such as classification and segmentation. We posit that this results from lack of explicit modelling of geometric information (e.g., locality in the signal or the orientation of a feature) in the latent space of CNFs. As such, we propose Equivariant Neural Fields (ENFs), a novel CNF architecture which uses a geometry-informed cross-attention to condition the NeF on a geometric variable--a latent point cloud of features--that enables an equivariant decoding from latent to field. We show that this approach induces a steerability property by which both field and latent are grounded in geometry and amenable to transformation laws: if the field transforms, the latent representation transforms accordingly--and vice versa. Crucially, this equivariance relation ensures that the latent is capable of (1) representing geometric patterns faithfully, allowing for geometric reasoning in latent space, and (2) weight-sharing over similar local patterns, allowing for efficient learning of datasets of fields. We validate these main properties in a range of tasks including classification, segmentation, forecasting, reconstruction and generative modelling, showing clear improvement over baselines with a geometry-free latent space. Code attached to submission https://github.com/Dafidofff/enf-jax. Code for a clean and minimal repo https://github.com/david-knigge/enf-min-jax.
format Preprint
id arxiv_https___arxiv_org_abs_2406_05753
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Grounding Continuous Representations in Geometry: Equivariant Neural Fields
Wessels, David R
Knigge, David M
Papa, Samuele
Valperga, Riccardo
Vadgama, Sharvaree
Gavves, Efstratios
Bekkers, Erik J
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Conditional Neural Fields (CNFs) are increasingly being leveraged as continuous signal representations, by associating each data-sample with a latent variable that conditions a shared backbone Neural Field (NeF) to reconstruct the sample. However, existing CNF architectures face limitations when using this latent downstream in tasks requiring fine-grained geometric reasoning, such as classification and segmentation. We posit that this results from lack of explicit modelling of geometric information (e.g., locality in the signal or the orientation of a feature) in the latent space of CNFs. As such, we propose Equivariant Neural Fields (ENFs), a novel CNF architecture which uses a geometry-informed cross-attention to condition the NeF on a geometric variable--a latent point cloud of features--that enables an equivariant decoding from latent to field. We show that this approach induces a steerability property by which both field and latent are grounded in geometry and amenable to transformation laws: if the field transforms, the latent representation transforms accordingly--and vice versa. Crucially, this equivariance relation ensures that the latent is capable of (1) representing geometric patterns faithfully, allowing for geometric reasoning in latent space, and (2) weight-sharing over similar local patterns, allowing for efficient learning of datasets of fields. We validate these main properties in a range of tasks including classification, segmentation, forecasting, reconstruction and generative modelling, showing clear improvement over baselines with a geometry-free latent space. Code attached to submission https://github.com/Dafidofff/enf-jax. Code for a clean and minimal repo https://github.com/david-knigge/enf-min-jax.
title Grounding Continuous Representations in Geometry: Equivariant Neural Fields
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2406.05753