Probabilistic Point Cloud Modeling via Self-Organizing Gaussian Mixture Models

TL;DR

This paper introduces a self-organizing Gaussian Mixture Model approach for probabilistic point cloud modeling that automatically adapts to scene complexity, improving generalization across diverse environments.

Contribution

It proposes a novel self-organizing GMM method that dynamically adjusts model complexity based on data, addressing limitations of existing fixed-parameter models.

Findings

Outperforms existing point cloud modeling techniques on real-world data.

Automatically adapts to scene complexity without manual tuning.

Demonstrates robustness across diverse environments.

Abstract

This letter presents a continuous probabilistic modeling methodology for spatial point cloud data using finite Gaussian Mixture Models (GMMs) where the number of components are adapted based on the scene complexity. Few hierarchical and adaptive methods have been proposed to address the challenge of balancing model fidelity with size. Instead, state-of-the-art mapping approaches require tuning parameters for specific use cases, but do not generalize across diverse environments. To address this gap, we utilize a self-organizing principle from information-theoretic learning to automatically adapt the complexity of the GMM model based on the relevant information in the sensor data. The approach is evaluated against existing point cloud modeling techniques on real-world data with varying degrees of scene complexity.