Irregular Convolutional Auto-Encoder on Point Clouds

TL;DR

This paper introduces a novel graph convolutional auto-encoder for point clouds that constructs sparse latent representations and enables detailed reconstruction and particle simulation acceleration.

Contribution

It presents a new non-isotropic convolution operation on irregular geometries and demonstrates its effectiveness in point cloud auto-encoding and fluid particle simulation.

Findings

Effective reconstruction of point clouds from sparse latent representations

Accelerated particle simulation using encoded latent clouds

Outperforms existing models on ShapeNetCore and fluid datasets

Abstract

We proposed a novel graph convolutional neural network that could construct a coarse, sparse latent point cloud from a dense, raw point cloud. With a novel non-isotropic convolution operation defined on irregular geometries, the model then can reconstruct the original point cloud from this latent cloud with fine details. Furthermore, we proposed that it is even possible to perform particle simulation using the latent cloud encoded from some simulated particle cloud (e.g. fluids), to accelerate the particle simulation process. Our model has been tested on ShapeNetCore dataset for Auto-Encoding with a limited latent dimension and tested on a synthesis dataset for fluids simulation. We also compare the model with other state-of-the-art models, and several visualizations were done to intuitively understand the model.