Learning Representations from 3D Gaussian Splats
For researchers in 3D scene understanding, this study provides a comparative analysis of representation learning on Gaussian Splats, but the findings are incremental and lack strong quantitative results.
This work evaluates geometric deep learning architectures for classifying 3D scenes represented via Gaussian Splatting, benchmarking point-based and graph-based models on point cloud and Gaussian Splatting datasets. Results show architectural families differ consistently and Gaussian-specific attributes affect representation quality.
3D Gaussian Splatting (3DGS) is a recent approach for scene rendering. Although primarily designed for view synthesis, its potential for scene understanding tasks remains underexplored. In this work, we conduct a comparative evaluation of various geometric deep learning architectures for the classification of 3D scenes represented using Gaussian Splatting. We benchmark point-based and graph-based models across both traditional point cloud datasets and dedicated Gaussian Splatting datasets. Scenes are embedded into latent representations, which are evaluated through end-to-end classification, linear probing, and clustering analysis. Our study provides insight into the suitability of different geometry-aware architectures and input feature configurations for learning effective 3D Gaussian Splat representations. The results highlight consistent differences between architectural families and reveal the impact of Gaussian-specific attributes on the quality of representation.