FPConv:学习点卷积的局部平坦化
FPConv: Learning Local Flattening for Point Convolution
摘要:
我们介绍了FPConv,一种新的表面风格的卷积运算符,设计用于3D点云分析。与以前的方法不同,FPConv不需要转换为3D网格或图形等中间表示,直接作用于点云的表面几何。更具体地说,对于每个点,FPConv通过自动学习权重图来执行局部平坦化,以将周围的点柔和地投影到2D网格上。因此,规则的2D卷积可以用于有效的特征学习。FPConv可以轻松集成到各种网络架构中,用于3D对象分类和3D场景分割等任务,并实现与现有体积类型卷积相当的性能。更重要的是,我们的实验还表明,FPConv可以作为体积卷积的补充,联合训练它们可以进一步将整体性能提升到最先进的结果。
英文摘要:
We introduce FPConv, a novel surface-style convolution operator designed for 3D point cloud analysis. Unlike previous methods, FPConv doesn't require transforming to intermediate representation like 3D grid or graph and directly works on surface geometry of point cloud. To be more specific, for each point, FPConv performs a local flattening by automatically learning a weight map to softly project surrounding points onto a 2D grid. Regular 2D convolution can thus be applied for efficient feature learning. FPConv can be easily integrated into various network architectures for tasks like 3D object classification and 3D scene segmentation, and achieve comparable performance with existing volumetric-type convolutions. More importantly, our experiments also show that FPConv can be a complementary of volumetric convolutions and jointly training them can further boost overall performance into state-of-the-art results.
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。