STITCH: Surface reconstrucTion using Implicit neural representations with Topology Constraints and persistent Homology

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jignasu, Anushrut, Herron, Ethan, Jiang, Zhanhong, Sarkar, Soumik, Hegde, Chinmay, Ganapathysubramanian, Baskar, Balu, Aditya, Krishnamurthy, Adarsh
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929667374579712
author Jignasu, Anushrut
Herron, Ethan
Jiang, Zhanhong
Sarkar, Soumik
Hegde, Chinmay
Ganapathysubramanian, Baskar
Balu, Aditya
Krishnamurthy, Adarsh
author_facet Jignasu, Anushrut
Herron, Ethan
Jiang, Zhanhong
Sarkar, Soumik
Hegde, Chinmay
Ganapathysubramanian, Baskar
Balu, Aditya
Krishnamurthy, Adarsh
contents We present STITCH, a novel approach for neural implicit surface reconstruction of a sparse and irregularly spaced point cloud while enforcing topological constraints (such as having a single connected component). We develop a new differentiable framework based on persistent homology to formulate topological loss terms that enforce the prior of a single 2-manifold object. Our method demonstrates excellent performance in preserving the topology of complex 3D geometries, evident through both visual and empirical comparisons. We supplement this with a theoretical analysis, and provably show that optimizing the loss with stochastic (sub)gradient descent leads to convergence and enables reconstructing shapes with a single connected component. Our approach showcases the integration of differentiable topological data analysis tools for implicit surface reconstruction.
format Preprint
id arxiv_https___arxiv_org_abs_2412_18696
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle STITCH: Surface reconstrucTion using Implicit neural representations with Topology Constraints and persistent Homology
Jignasu, Anushrut
Herron, Ethan
Jiang, Zhanhong
Sarkar, Soumik
Hegde, Chinmay
Ganapathysubramanian, Baskar
Balu, Aditya
Krishnamurthy, Adarsh
Computer Vision and Pattern Recognition
Graphics
Machine Learning
We present STITCH, a novel approach for neural implicit surface reconstruction of a sparse and irregularly spaced point cloud while enforcing topological constraints (such as having a single connected component). We develop a new differentiable framework based on persistent homology to formulate topological loss terms that enforce the prior of a single 2-manifold object. Our method demonstrates excellent performance in preserving the topology of complex 3D geometries, evident through both visual and empirical comparisons. We supplement this with a theoretical analysis, and provably show that optimizing the loss with stochastic (sub)gradient descent leads to convergence and enables reconstructing shapes with a single connected component. Our approach showcases the integration of differentiable topological data analysis tools for implicit surface reconstruction.
title STITCH: Surface reconstrucTion using Implicit neural representations with Topology Constraints and persistent Homology
topic Computer Vision and Pattern Recognition
Graphics
Machine Learning
url https://arxiv.org/abs/2412.18696