Probabilistic Point Cloud Modeling via Self-Organizing Gaussian Mixture Models

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 scene complexity. Few hierarchical and adaptive methods have been proposed to address challenge balancing model fidelity with size. Instead, state-of-the-art mapping approaches require tuning parameters specific use cases, but do not generalize across diverse environments. To this gap, we utilize self-organizing principle from information-theoretic learning automatically adapt complexity GMM relevant information in sensor data. The approach is evaluated against existing techniques real-world varying degrees