DeepJoin: Learning a Joint Occupancy, Signed Distance, and Normal Field Function for Shape Repair

TL;DR

DeepJoin is a neural network-based method that generates high-resolution, artifact-free repairs for fractured 3D shapes by inferring complete shapes and break surfaces using a novel implicit shape representation.

Contribution

It introduces a new implicit shape representation combining occupancy, signed distance, and normal fields for shape repair, enabling high-resolution and artifact-free restorations.

Findings

Outperforms baseline methods in chamfer distance and normal consistency

Works on synthetic, scanned, and cultural heritage 3D objects

Produces surface-continuous restorations without artifacts

Abstract

We introduce DeepJoin, an automated approach to generate high-resolution repairs for fractured shapes using deep neural networks. Existing approaches to perform automated shape repair operate exclusively on symmetric objects, require a complete proxy shape, or predict restoration shapes using low-resolution voxels which are too coarse for physical repair. We generate a high-resolution restoration shape by inferring a corresponding complete shape and a break surface from an input fractured shape. We present a novel implicit shape representation for fractured shape repair that combines the occupancy function, signed distance function, and normal field. We demonstrate repairs using our approach for synthetically fractured objects from ShapeNet, 3D scans from the Google Scanned Objects dataset, objects in the style of ancient Greek pottery from the QP Cultural Heritage dataset, and real fractured objects. We outperform three baseline approaches in terms of chamfer distance and normal consistency. Unlike existing approaches and restorations using subtraction, DeepJoin restorations do not exhibit surface artifacts and join closely to the fractured region of the fractured shape. Our code is available at: https://github.com/Terascale-All-sensing-Research-Studio/DeepJoin.