arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.07358cs.CVeess.IV
收藏

SoftPoolNet:用于点云补全和分类的形状描述符

SoftPoolNet: Shape Descriptor for Point Cloud Completion and Classification

Technische Universität München(慕尼黑工业大学);Google Inc.(谷歌公司)

Yida Wang, David Joseph Tan, Nassir Navab, Federico Tombari

摘要:

点云通常是许多应用程序的默认选择,因为它们比体积数据更具灵活性和效率。然而,它们的无组织性-点以无序的方式存储-使它们不太适合由深度学习管道处理。本文提出了一种基于点云的三维物体完成和分类方法。我们引入了一种新的方式来组织提取的功能的基础上,他们的激活,我们命名为软池。对于解码器阶段,我们提出了区域卷积,一种旨在最大化全局激活熵的新算子。此外,受点完备网络(PCN)中局部细化过程的启发,我们还提出了一种分片变形操作来模拟点云的解卷积操作。本文证明了我们的区域激活可以被纳入许多点云架构,如PakasNet和PCN,从而获得更好的几何完成性能。我们在不同的3D任务(如对象完成和分类)上评估我们的方法,以实现最先进的精度。

英文摘要:

Point clouds are often the default choice for many applications as they exhibit more flexibility and efficiency than volumetric data. Nevertheless, their unorganized nature -- points are stored in an unordered way -- makes them less suited to be processed by deep learning pipelines. In this paper, we propose a method for 3D object completion and classification based on point clouds. We introduce a new way of organizing the extracted features based on their activations, which we name soft pooling. For the decoder stage, we propose regional convolutions, a novel operator aimed at maximizing the global activation entropy. Furthermore, inspired by the local refining procedure in Point Completion Network (PCN), we also propose a patch-deforming operation to simulate deconvolutional operations for point clouds. This paper proves that our regional activation can be incorporated in many point cloud architectures like AtlasNet and PCN, leading to better performance for geometric completion. We evaluate our approach on different 3D tasks such as object completion and classification, achieving state-of-the-art accuracy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑