Structure preserving implicit shape encoding via flow regularization

Sven Dummer; Nicola Strisciuglio; Christoph Brune

Structure preserving implicit shape encoding via flow regularization

Sven Dummer, Nicola Strisciuglio, Christoph Brune

Research output: Contribution to conference › Abstract › Academic

7 Downloads (Pure)

Abstract

Recently, implicit neural representations (INRs) emerged as an effective method for reconstructing shapes. Several of such methods transform templates to target shapes. Current template-based methods lack proper regularization. In this work, we add a novel regularization to a deformable template approach and discuss the benefits of this regularization with a simple test case.

Original language	English
Number of pages	4
Publication status	Published - 18 Nov 2022
Event	Geometric Deep Learning in Medical Image Analysis - Hotel CASA, Amsterdam, Netherlands Duration: 18 Dec 2021 → 18 Dec 2021

Conference

Conference	Geometric Deep Learning in Medical Image Analysis
Abbreviated title	GeoMedIA
Country/Territory	Netherlands
City	Amsterdam
Period	18/12/21 → 18/12/21

Keywords

Latent space
Shape reconstruction
Implicit neural representation
Neural ODE

Access to Document

GeoMedia_poster_Sven_DummerFinal published version, 2.82 MB

https://openreview.net/pdf?id=YcjlgyX_Ur1

Normal Form Autoencoder, data associated with the publication: ‘Learning normal form autoencoders for data-driven discovery of universal, parameter-dependent governing equations’
Kalia, M. (Creator), Meijer, H. G. E. (Creator), Brunton, S. L. (Creator), Kutz, J. N. (Creator) & Brune, C. (Creator), 4TU.Centre for Research Data, 18 Jun 2021
Dataset

Cite this

@conference{405026006a5c4163ade33aea23af93c3,

title = "Structure preserving implicit shape encoding via flow regularization",

abstract = "Recently, implicit neural representations (INRs) emerged as an effective method for reconstructing shapes. Several of such methods transform templates to target shapes. Current template-based methods lack proper regularization. In this work, we add a novel regularization to a deformable template approach and discuss the benefits of this regularization with a simple test case.",

keywords = "Latent space, Shape reconstruction, Implicit neural representation, Neural ODE",

author = "Sven Dummer and Nicola Strisciuglio and Christoph Brune",

year = "2022",

month = nov,

day = "18",

language = "English",

note = "Geometric Deep Learning in Medical Image Analysis, GeoMedIA ; Conference date: 18-12-2021 Through 18-12-2021",

}

TY - CONF

T1 - Structure preserving implicit shape encoding via flow regularization

AU - Dummer, Sven

AU - Strisciuglio, Nicola

AU - Brune, Christoph

PY - 2022/11/18

Y1 - 2022/11/18

N2 - Recently, implicit neural representations (INRs) emerged as an effective method for reconstructing shapes. Several of such methods transform templates to target shapes. Current template-based methods lack proper regularization. In this work, we add a novel regularization to a deformable template approach and discuss the benefits of this regularization with a simple test case.

AB - Recently, implicit neural representations (INRs) emerged as an effective method for reconstructing shapes. Several of such methods transform templates to target shapes. Current template-based methods lack proper regularization. In this work, we add a novel regularization to a deformable template approach and discuss the benefits of this regularization with a simple test case.

KW - Latent space

KW - Shape reconstruction

KW - Implicit neural representation

KW - Neural ODE

M3 - Abstract

T2 - Geometric Deep Learning in Medical Image Analysis

Y2 - 18 December 2021 through 18 December 2021

ER -

Structure preserving implicit shape encoding via flow regularization

Abstract

Conference

Keywords

Access to Document

Datasets

Normal Form Autoencoder, data associated with the publication: ‘Learning normal form autoencoders for data-driven discovery of universal, parameter-dependent governing equations’

Cite this