Benchmarking Robustness of 3D Point Cloud Recognition Against Common Corruptions

Sample Visualizations from our ModelNet40-C Dataset. Our dataset contains 185,100 point clouds from 40 classes, 15 corruption types, and 5 severity levels. We provide a detailed taxonomy of the constructed corruption types. ModelNet40-C is, to the best of our knowledge, the first comprehensive dataset for benchmarking corruption robustness of 3D point cloud classification.

Deep neural networks on 3D point cloud data have been widely used in the real world, especially in safety-critical applications. However, their robustness against corruptions is less studied. In this paper, we present ModelNet40-C, the first comprehensive benchmark on 3D point cloud corruption robustness, consisting of 15 common and realistic corruptions. Our evaluation shows a significant gap between the performances on ModelNet40 and ModelNet40-C for state-of-the-art (SOTA) models. To reduce the gap, we propose a simple but effective method by combining PointCutMix-R and TENT after evaluating a wide range of augmentation and test-time adaptation strategies. We identify a number of critical insights for future studies on corruption robustness in point cloud recognition. For instance, we unveil that Transformer-based architectures with proper training recipes achieve the strongest robustness. We hope our in-depth analysis will motivate the development of robust training strategies or architecture designs in the 3D point cloud domain.

@article{sun2022benchmarking,

title={Benchmarking Robustness of 3D Point Cloud Recognition Against Common Corruptions},

author={Jiachen Sun and Qingzhao Zhang and Bhavya Kailkhura and Zhiding Yu and Chaowei Xiao and Z. Morley Mao},

year={2022},

journal={arXiv preprint arXiv:2201.12296},

primaryClass={cs.LG}

}