We introduce ModelNet40-E, a new benchmark designed to assess the robustness and calibration of point cloud classification models under synthetic LiDAR-like noise. Unlike existing benchmarks, ModelNet40-E provides both noise-corrupted point clouds and point-wise uncertainty annotations via Gaussian noise parameters ({\sigma}, {\mu}), enabling fine-grained evaluation of uncertainty modeling. We evaluate three popular models-PointNet, DGCNN, and Point Transformer v3-across multiple noise levels using classification accuracy, calibration metrics, and uncertainty-awareness. While all models degrade under increasing noise, Point Transformer v3 demonstrates superior calibration, with predicted uncertainties more closely aligned with the underlying measurement uncertainty.
@article{arxiv.2508.01269,
title = {ModelNet40-E: An Uncertainty-Aware Benchmark for Point Cloud Classification},
author = {Pedro Alonso and Tianrui Li and Chongshou Li},
journal= {arXiv preprint arXiv:2508.01269},
year = {2025}
}