Local Latent Representation based on Geometric Convolution for Particle Data Feature Exploration

TL;DR

This paper introduces a geometric convolution-based neural network approach to transform particle data into latent representations, enabling feature extraction and tracking in scientific applications without relying on hand-crafted descriptors.

Contribution

It proposes a novel method using geometric convolution for 3D particle data, capturing local features in a learned latent space for improved analysis.

Findings

Latent representations effectively capture particle positions and attributes.

Features extracted are comparable to traditional hand-crafted methods.

Tracking results demonstrate the method's applicability across datasets.

Abstract

Feature related particle data analysis plays an important role in many scientific applications such as fluid simulations, cosmology simulations and molecular dynamics. Compared to conventional methods that use hand-crafted feature descriptors, some recent studies focus on transforming the data into a new latent space, where features are easier to be identified, compared and extracted. However, it is challenging to transform particle data into latent representations, since the convolution neural networks used in prior studies require the data presented in regular grids. In this paper, we adopt Geometric Convolution, a neural network building block designed for 3D point clouds, to create latent representations for scientific particle data. These latent representations capture both the particle positions and their physical attributes in the local neighborhood so that features can be extracted by clustering in the latent space, and tracked by applying tracking algorithms such as mean-shift. We validate the extracted features and tracking results from our approach using datasets from three applications and show that they are comparable to the methods that define hand-crafted features for each specific dataset.