Mesh Processing Non-Meshes via Neural Displacement Fields

2025-08-16

AI Summary

Traditional mesh processing pipelines struggle to accommodate emerging non-mesh surface representations—such as neural implicit fields—due to their reliance on costly explicit meshing or transmission of large mesh datasets, undermining streaming efficiency. This paper introduces a neural displacement field framework that learns a differentiable mapping from a coarse input mesh to a neural implicit field, enabling manifold-conforming mesh extraction and scalar field compression without high-fidelity mesh reconstruction. Our method integrates sparse voxel sampling with differentiable rendering to jointly optimize geometric fidelity, computational efficiency, and storage compactness. Experiments demonstrate that complex geometries and associated attributes can be compressed to hundreds of kilobytes, supporting millisecond-scale mesh extraction, real-time interactive editing, and intrinsic shape analysis. The approach achieves a significant balance among reconstruction accuracy, processing speed, and bandwidth efficiency.

Technology Category

• Machine Learning: Learning with Manifolds
• Data Mining & Knowledge Management: Data Compression
• Computer Vision: 3D Computer Vision

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Mesh processing pipelines are mature, but adapting them to newer non-mesh surface representations -- which enable fast rendering with compact file size -- requires costly meshing or transmitting bulky meshes, negating their core benefits for streaming applications. We present a compact neural field that enables common geometry processing tasks across diverse surface representations. Given an input surface, our method learns a neural map from its coarse mesh approximation to the surface. The full representation totals only a few hundred kilobytes, making it ideal for lightweight transmission. Our method enables fast extraction of manifold and Delaunay meshes for intrinsic shape analysis, and compresses scalar fields for efficient delivery of costly precomputed results. Experiments and applications show that our fast, compact, and accurate approach opens up new possibilities for interactive geometry processing.

Problem

• Adapting mesh pipelines to non-mesh surfaces efficiently
• Enabling geometry processing across diverse surface representations
• Compressing scalar fields for efficient data delivery

Innovation

• Neural map from coarse mesh to surface
• Compact neural field for diverse surfaces
• Fast extraction of manifold meshes