论文阅读【CVPR-2022】3D Common Corruptions and Data Augmentation

3D Common Corruptions and Data Augmentation

3维通用Corruptions和数据扩增

studyai.com

搜索论文: 3D Common Corruptions and Data Augmentation

摘要(Abstract)

We introduce a set of image transformations that can be used as ‘corruptions’ to evaluate the robustness of models as well as ‘data augmentation’ mechanisms for training neural networks.

我们引入了一组图像变换，可被用作 corruptions 来评估模型的鲁棒性以及用于神经网络训练中的数据扩增机制。

The primary distinction of the proposed transformations is that, unlike existing approaches such as Common Corruptions [27], the geometry of the scene is incorporated in the transformations – thus leading to corruptions that are more likely to occur in the real world.

我们提出的这组图像变换与已有的变换(如文献27的通用Corruptions )的主要区别是：场景的几何特征也被包含在变换之中，因此产生的corruptions与真实世界中的corruptions更加相似。

We show these transformations are ‘efficient’ (can be computed on-the-fly), ‘extendable’ (can be applied on most datasets of real images), expose vulnerability of existing models, and can effectively make models more robust when employed as ‘3D data augmentation’ mechanisms.

我们展示了这些变换的很多优点：‘高效的’(可以在运行阶段计算)，’可扩展的’(可被用于真实图像构成的很多数据集上)，可以暴露现有模型的弱点，当被用于3D数据扩增机制的时候可以有效的使得模型更加鲁棒。

Our evaluations performed on several tasks and datasets suggest incorporating 3D information into robustness benchmarking and training opens up a promising direction for robustness research.

我们对若干个任务和数据集进行的评估表明，将3D信息纳入鲁棒性基准测试和训练过程可以为鲁棒性研究开辟了一个有前景的方向