Revisiting Point Cloud Simplification: A Learnable Feature Preserving Approach

TL;DR

This paper introduces a fast, learnable point cloud simplification method using graph neural networks that preserves salient features and generalizes well to new shapes, reducing computational costs.

Contribution

A novel graph neural network-based approach for point cloud simplification that is fast, learnable, and capable of zero-shot generalization to unseen shapes.

Findings

Effective preservation of salient features in simplified point clouds

Outperforms traditional methods in speed and quality

Demonstrates strong generalization to out-of-distribution shapes

Abstract

The recent advances in 3D sensing technology have made possible the capture of point clouds in significantly high resolution. However, increased detail usually comes at the expense of high storage, as well as computational costs in terms of processing and visualization operations. Mesh and Point Cloud simplification methods aim to reduce the complexity of 3D models while retaining visual quality and relevant salient features. Traditional simplification techniques usually rely on solving a time-consuming optimization problem, hence they are impractical for large-scale datasets. In an attempt to alleviate this computational burden, we propose a fast point cloud simplification method by learning to sample salient points. The proposed method relies on a graph neural network architecture trained to select an arbitrary, user-defined, number of points from the input space and to re-arrange their positions so as to minimize the visual perception error. The approach is extensively evaluated on various datasets using several perceptual metrics. Importantly, our method is able to generalize to out-of-distribution shapes, hence demonstrating zero-shot capabilities.