PriFit: Learning to Fit Primitives Improves Few Shot Point Cloud Segmentation. (6th October 2022)

• Record Type: Journal Article
• Authors: Sharma, G.; Dash, B.; RoyChowdhury, A.; Gadelha, M.; Loizou, M.; Cao, L.; Wang, R.; Learned‐Miller, E. G.; Maji, S.; Kalogerakis, E.
• Is Part Of: Computer graphics forum. Volume 41:Number 5(2022)
• Page Start: 39
• Page End: 50
• Publication Date: 2022-10-06
• Subjects: CCS Concepts -- Computing methodologies → Shape representations -- Neural networks -- Theory of computation → Semi‐supervised learning; Computer graphics -- Periodicals; 006.605
• DOI: 10.1111/cgf.14601
• Languages: English
• ISSNs: 0167-7055

Abstract

Abstract: We present Pri Fit, a semi‐supervised approach for label‐efficient learning of 3D point cloud segmentation networks. Pri Fit combines geometric primitive fitting with point‐based representation learning. Its key idea is to learn point representations whose clustering reveals shape regions that can be approximated well by basic geometric primitives, such as cuboids and ellipsoids. The learned point representations can then be re‐used in existing network architectures for 3D point cloud segmentation, and improves their performance in the few‐shot setting. According to our experiments on the widely used ShapeNet and PartNet benchmarks, Pri Fit outperforms several state‐of‐the‐art methods in this setting, suggesting that decomposability into primitives is a useful prior for learning representations predictive of semantic parts. We present a number of ablative experiments varying the choice of geometric primitives and downstream tasks to demonstrate the effectiveness of the method.