Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Adaptive Scaling with Geometric and Visual Continuity of completed 3D objects

Jelle Vermandere, Maarten Bassier, Maarten Vergauwen

TL;DR

This work introduces a part-aware scaling framework that transforms static completed SDFs into editable, structurally coherent objects and incorporates a repetition-based strategy to handle large-scale deformations while preserving repeating geometric patterns.

Abstract

Object completion networks typically produce static Signed Distance Fields (SDFs) that faithfully reconstruct geometry but cannot be rescaled or deformed without introducing structural distortions. This limitation restricts their use in applications requiring flexible object manipulation, such as indoor redesign, simulation, and digital content creation. We introduce a part-aware scaling framework that transforms these static completed SDFs into editable, structurally coherent objects. Starting from SDFs and Texture Fields generated by state-of-the-art completion models, our method performs automatic part segmentation, defines user-controlled scaling zones, and applies smooth interpolation of SDFs, color, and part indices to enable proportional and artifact-free deformation. We further incorporate a repetition-based strategy to handle large-scale deformations while preserving repeating geometric patterns. Experiments on Matterport3D and ShapeNet objects show that our method overcomes the inherent rigidity of completed SDFs and is visually more appealing than global and naive selective scaling, particularly for complex shapes and repetitive structures.

Adaptive Scaling with Geometric and Visual Continuity of completed 3D objects

TL;DR

This work introduces a part-aware scaling framework that transforms static completed SDFs into editable, structurally coherent objects and incorporates a repetition-based strategy to handle large-scale deformations while preserving repeating geometric patterns.

Abstract

Object completion networks typically produce static Signed Distance Fields (SDFs) that faithfully reconstruct geometry but cannot be rescaled or deformed without introducing structural distortions. This limitation restricts their use in applications requiring flexible object manipulation, such as indoor redesign, simulation, and digital content creation. We introduce a part-aware scaling framework that transforms these static completed SDFs into editable, structurally coherent objects. Starting from SDFs and Texture Fields generated by state-of-the-art completion models, our method performs automatic part segmentation, defines user-controlled scaling zones, and applies smooth interpolation of SDFs, color, and part indices to enable proportional and artifact-free deformation. We further incorporate a repetition-based strategy to handle large-scale deformations while preserving repeating geometric patterns. Experiments on Matterport3D and ShapeNet objects show that our method overcomes the inherent rigidity of completed SDFs and is visually more appealing than global and naive selective scaling, particularly for complex shapes and repetitive structures.
Paper Structure (20 sections, 3 equations, 6 figures)

This paper contains 20 sections, 3 equations, 6 figures.

Table of Contents

  1. Introduction
  2. Background and related work
  3. Object Completion
  4. Part Segmentation
  5. Surface-Based Deformation
  6. SDF-Based Deformation
  7. Texture Deformation
  8. Methodology
  9. Preprocessing: Generating CSDFs
  10. Part Segmentation
  11. Scaling Zone Definition
  12. SDF, Color, and Part Index Interpolation
  13. Linear Interpolation of SDF
  14. Linear Interpolation of Colour
  15. Handling Part Index Interpolation
  16. ...and 5 more sections

Figures (6)

  • Figure 1: Overview of the proposed pipeline, starting with the object completion (left), going into the part segmentation (centre-left), followed scaling zone definition (centre-right) to result in a scaled and coloured object (right).
  • Figure 2: The CSDF stored as a 3D texture, sliced in an 8x8 grid (left) and the rendered object (right)
  • Figure 3: The CSDF rendered as a 3D texture(left) and the convex decomposition of the object (right).
  • Figure 4: The object with a small repeatable part selected(left) and the resulting repeated parts in the scaled object (right).
  • Figure 5: The results of the object scaling pipeline. top: The matterport objects, bottom: the Sketchfab objects.
  • ...and 1 more figures