Point Cloud Denoising via Momentum Ascent in Gradient Fields

TL;DR

This paper introduces a momentum gradient ascent method for point cloud denoising that enhances stability and efficiency by leveraging previous iteration information, outperforming existing methods across various noise conditions.

Contribution

The paper proposes a novel momentum-based gradient ascent approach that improves stability and reduces inference time in point cloud denoising compared to prior gradient-based methods.

Findings

Outperforms state-of-the-art denoising methods

Provides more stable and faster denoising results

Effective across different noise types and levels

Abstract

To achieve point cloud denoising, traditional methods heavily rely on geometric priors, and most learning-based approaches suffer from outliers and loss of details. Recently, the gradient-based method was proposed to estimate the gradient fields from the noisy point clouds using neural networks, and refine the position of each point according to the estimated gradient. However, the predicted gradient could fluctuate, leading to perturbed and unstable solutions, as well as a long inference time. To address these issues, we develop the momentum gradient ascent method that leverages the information of previous iterations in determining the trajectories of the points, thus improving the stability of the solution and reducing the inference time. Experiments demonstrate that the proposed method outperforms state-of-the-art approaches with a variety of point clouds, noise types, and noise levels. Code is available at: https://github.com/IndigoPurple/MAG