Regularization Strategy for Point Cloud via Rigidly Mixed Sample
This work addresses the lack of effective data augmentation strategies for point cloud processing, which is crucial for improving the generalization of deep neural networks on small point cloud datasets.
This paper introduces Rigid Subset Mix (RSMix), a new data augmentation technique for point clouds that creates virtual mixed samples by combining shape-preserved subsets from different samples. RSMix significantly improved shape classification performance by regularizing deep neural networks.
Data augmentation is an effective regularization strategy to alleviate the overfitting, which is an inherent drawback of the deep neural networks. However, data augmentation is rarely considered for point cloud processing despite many studies proposing various augmentation methods for image data. Actually, regularization is essential for point clouds since lack of generality is more likely to occur in point cloud due to small datasets. This paper proposes a Rigid Subset Mix (RSMix), a novel data augmentation method for point clouds that generates a virtual mixed sample by replacing part of the sample with shape-preserved subsets from another sample. RSMix preserves structural information of the point cloud sample by extracting subsets from each sample without deformation using a neighboring function. The neighboring function was carefully designed considering unique properties of point cloud, unordered structure and non-grid. Experiments verified that RSMix successfully regularized the deep neural networks with remarkable improvement for shape classification. We also analyzed various combinations of data augmentations including RSMix with single and multi-view evaluations, based on abundant ablation studies.