Phase Transitions, Distance Functions, and Implicit Neural Representations

06/14/2021
by   Yaron Lipman, et al.
0

Representing surfaces as zero level sets of neural networks recently emerged as a powerful modeling paradigm, named Implicit Neural Representations (INRs), serving numerous downstream applications in geometric deep learning and 3D vision. Training INRs previously required choosing between occupancy and distance function representation and different losses with unknown limit behavior and/or bias. In this paper we draw inspiration from the theory of phase transitions of fluids and suggest a loss for training INRs that learns a density function that converges to a proper occupancy function, while its log transform converges to a distance function. Furthermore, we analyze the limit minimizer of this loss showing it satisfies the reconstruction constraints and has minimal surface perimeter, a desirable inductive bias for surface reconstruction. Training INRs with this new loss leads to state-of-the-art reconstructions on a standard benchmark.

READ FULL TEXT

page 2

page 3

page 5

page 6

page 10

page 12

page 14

page 17

research
07/20/2020

Coupling Explicit and Implicit Surface Representations for Generative 3D Modeling

We propose a novel neural architecture for representing 3D surfaces, whi...
research
04/16/2021

Signed Distance Function Computation from an Implicit Surface

We describe in this short note a technique to convert an implicit surfac...
research
03/25/2023

VisCo Grids: Surface Reconstruction with Viscosity and Coarea Grids

Surface reconstruction has been seeing a lot of progress lately by utili...
research
12/11/2020

Iso-Points: Optimizing Neural Implicit Surfaces with Hybrid Representations

Neural implicit functions have emerged as a powerful representation for ...
research
05/28/2023

StEik: Stabilizing the Optimization of Neural Signed Distance Functions and Finer Shape Representation

We present new insights and a novel paradigm (StEik) for learning implic...
research
03/17/2023

Fluid Dynamics Network: Topology-Agnostic 4D Reconstruction via Fluid Dynamics Priors

Representing 3D surfaces as level sets of continuous functions over ℝ^3 ...
research
02/12/2021

Applicability of Random Matrix Theory in Deep Learning

We investigate the local spectral statistics of the loss surface Hessian...

Please sign up or login with your details

Forgot password? Click here to reset