A 3D mesh convolution-based autoencoder for geometry compression

TL;DR

This paper presents a new 3D mesh convolution autoencoder that effectively compresses and reconstructs irregular mesh geometries without preprocessing, outperforming existing methods in accuracy and classification.

Contribution

It introduces a novel mesh convolution autoencoder that handles irregular meshes directly, with dedicated pooling operations and a consistent latent space, advancing geometry compression techniques.

Findings

Outperforms state-of-the-art in geometry reconstruction

Achieves better latent space classification results

Handles irregular meshes without preprocessing

Abstract

In this paper, we introduce a novel 3D mesh convolution-based autoencoder for geometry compression, able to deal with irregular mesh data without requiring neither preprocessing nor manifold/watertightness conditions. The proposed approach extracts meaningful latent representations by learning features directly from the mesh faces, while preserving connectivity through dedicated pooling and unpooling operations. The encoder compresses the input mesh into a compact base mesh space, which ensures that the latent space remains comparable. The decoder reconstructs the original connectivity and restores the compressed geometry to its full resolution. Extensive experiments on multi-class datasets demonstrate that our method outperforms state-of-the-art approaches in both 3D mesh geometry reconstruction and latent space classification tasks. Code available at: github.com/germainGB/MeshConv3D