Unsupervised contrastive learning with simple transformation for 3D point cloud data

Authors: Jiang, J., Lu, X., Ouyang, W. and Wang, M.

Journal: Visual Computer

ISSN: 0178-2789

DOI: 10.1007/s00371-023-02921-y

Abstract:

Though a number of point cloud learning methods have been proposed to handle unordered points, most of them are supervised and require labels for training. By contrast, unsupervised learning of point cloud data has received much less attention to date. In this paper, we propose a simple yet effective approach for unsupervised point cloud learning. In particular, we identify a very useful transformation which generates a good contrastive version of an original point cloud. They make up a pair. After going through a shared encoder and a shared head network, the consistency between the output representations are maximized with introducing two variants of contrastive losses to respectively facilitate downstream classification and segmentation. To demonstrate the efficacy of our method, we conduct experiments on three downstream tasks which are 3D object classification (on ModelNet40 and ModelNet10), shape part segmentation (on ShapeNet Part dataset) as well as scene segmentation (on S3DIS). Comprehensive results show that our unsupervised contrastive representation learning enables impressive outcomes in object classification and semantic segmentation. It generally outperforms current unsupervised methods, and even achieves comparable performance to supervised methods.

Source: Scopus

Unsupervised contrastive learning with simple transformation for 3D point cloud data

Authors: Jiang, J., Lu, X., Ouyang, W. and Wang, M.

Journal: VISUAL COMPUTER

eISSN: 1432-2315

ISSN: 0178-2789

DOI: 10.1007/s00371-023-02921-y

Source: Web of Science (Lite)

Unsupervised contrastive learning with simple transformation for 3D point cloud data

Authors: Jiang, J., Lu, X., Ouyang, W. and Wang, M.

Journal: The Visual Computer

Publisher: Springer Nature

ISSN: 0178-2789

DOI: 10.1007/s00371-023-02921-y

Abstract:

Though a number of point cloud learning methods have been proposed to handle unordered points, most of them are supervised and require labels for training. By contrast, unsupervised learning of point cloud data has received much less attention to date. In this paper, we propose a simple yet effective approach for unsupervised point cloud learning. In particular, we identify a very useful transformation which generates a good contrastive version of an original point cloud. They make up a pair. After going through a shared encoder and a shared head network, the consistency between the output representations are maximized with introducing two variants of contrastive losses to respectively facilitate downstream classification and segmentation. To demonstrate the efficacy of our method, we conduct experiments on three downstream tasks which are 3D object classification (on ModelNet40 and ModelNet10), shape part segmentation (on ShapeNet Part dataset) as well as scene segmentation (on S3DIS). Comprehensive results show that our unsupervised contrastive representation learning enables impressive outcomes in object classification and semantic segmentation. It generally outperforms current unsupervised methods, and even achieves comparable performance to supervised methods.

Source: Manual