ModelNet40-C

displayName: ModelNet40-C (ModelNet-C)
labelTypes: []
license:
- BSD 3-Clause
mediaTypes:
- PointCloud
paperUrl: https://arxiv.org/pdf/2201.12296v1.pdf
publishDate: "2022-01-28"
publishUrl: https://sites.google.com/umich.edu/modelnet40c
publisher:
- University of Michigan
- NVIDIA
- Arizona State University
- Lawrence Livermore National Laboratory
tags: []
taskTypes:
- 3D Point Cloud Classification
- Robust Classification
- 3D Classification

---
## 简介
ModelNet40-C 是一个综合数据集，用于对 3D 点云识别的损坏鲁棒性进行基准测试。
我们基于 ModelNet40 验证集创建 ModelNet40-C，该验证集具有 15 种损坏类型和每种损坏类型的 5 个严重级别，包括密度、噪声和转换损坏模式。我们的数据集包含 185,000 个不同的点云，有助于全面了解模型的稳健性。
## 引文
```
@article{sun2022benchmarking,
title={Benchmarking robustness of 3d point cloud recognition against common corruptions},
author={Sun, Jiachen and Zhang, Qingzhao and Kailkhura, Bhavya and Yu, Zhiding and Xiao, Chaowei and Mao, Z Morley},
journal={arXiv preprint arXiv:2201.12296},
year={2022}
}
```

## Download dataset
:modelscope-code[]{type="git"}

displayName: ModelNet40-C（ModelNet-C）
labelTypes: []
license:
- BSD 3条款许可证（BSD 3-Clause）
mediaTypes:
- 点云（PointCloud）
paperUrl: https://arxiv.org/pdf/2201.12296v1.pdf
publishDate: "2022-01-28"
publishUrl: https://sites.google.com/umich.edu/modelnet40c
publisher:
- 密歇根大学
- 英伟达（NVIDIA）
- 亚利桑那州立大学
- 劳伦斯利弗莫尔国家实验室
tags: []
taskTypes:
- 3D点云分类（3D Point Cloud Classification）
- 鲁棒分类
- 3D分类

---
## 简介
ModelNet40-C 是一款用于基准测试3D点云识别模型抗损坏鲁棒性的综合性数据集。本数据集基于ModelNet40验证集构建，涵盖15类损坏类型及每类对应的5种严重程度，包含密度扰动、噪声污染、空间变换等多种损坏模式。数据集总计包含185,000组独立点云，可用于全面评估模型的稳健性能。
## 引文

@article{sun2022benchmarking,
title={Benchmarking robustness of 3d point cloud recognition against common corruptions},
author={Sun, Jiachen and Zhang, Qingzhao and Kailkhura, Bhavya and Yu, Zhiding and Xiao, Chaowei and Mao, Z Morley},
journal={arXiv preprint arXiv:2201.12296},
year={2022}
}

## Download dataset
:modelscope-code[]{type="git"}