Point Cloud Augmentation with Weighted Local Transformations

Despite the extensive usage of point clouds in 3D vision, relatively limited data are available for training deep neural networks. Although data augmentation is a standard approach to compensate for the scarcity of data, it has been less explored in the point cloud literature. In this paper, we prop...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Proceedings / IEEE International Conference on Computer Vision s. 528 - 537
Hlavní autori: Kim, Sihyeon, Lee, Sanghyeok, Hwang, Dasol, Lee, Jaewon, Hwang, Seong Jae, Kim, Hyunwoo J.
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 01.10.2021
Predmet:
grouping and shape
Manuals
Neural networks
Point cloud compression
Recognition and classification
Segmentation
Shape
Task analysis
Three-dimensional displays
Training
Vision applications and systems
ISSN:2380-7504
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Despite the extensive usage of point clouds in 3D vision, relatively limited data are available for training deep neural networks. Although data augmentation is a standard approach to compensate for the scarcity of data, it has been less explored in the point cloud literature. In this paper, we propose a simple and effective augmentation method called PointWOLF for point cloud augmentation. The proposed method produces smoothly varying non-rigid deformations by locally weighted transformations centered at multiple anchor points. The smooth deformations allow diverse and realistic augmentations. Furthermore, in order to minimize the manual efforts to search the optimal hyperparameters for augmentation, we present AugTune, which generates augmented samples of desired difficulties producing targeted confidence scores. Our experiments show our framework consistently improves the performance for both shape classification and part segmentation tasks. Particularly, with PointNet++, PointWOLF achieves the state-of-the-art 89.7 accuracy on shape classification with the real-world ScanObjectNN dataset. The code is available at https://github.com/mlvlab/PointWOLF.
ISSN:2380-7504
DOI:10.1109/ICCV48922.2021.00059