{"ID":2887561,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.01269","arxiv_id":"2508.01269","title":"ModelNet40-E: An Uncertainty-Aware Benchmark for Point Cloud Classification","abstract":"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 (σ, μ), 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.","short_abstract":"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 (σ,...","url_abs":"https://arxiv.org/abs/2508.01269","url_pdf":"https://arxiv.org/pdf/2508.01269v2","authors":"[\"Pedro Alonso\",\"Tianrui Li\",\"Chongshou Li\"]","published":"2025-08-02T08:57:20Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Transformer\",\"Convolutional Neural Network\"]","has_code":false}