Dynamic Graph CNN for Learning on Point Clouds
Dynamic Graph CNN for Learning on Point Clouds
复制标题
DOI：
10.1145/3326362
复制
发表时间：
2019-11-01
影响因子：
6.2
通讯作者：
Solomon, Justin M.
中科院分区：
文献类型：
--
作者：
Wang, Yue;Sun, Yongbin;Solomon, Justin M.
点云提供了一种灵活的几何表示，适用于计算机图形学中的无数应用;它们还包括大多数3D数据采集设备的原始输出。虽然在图形和视觉领域早就提出了点云上的手工设计特征，但是最近卷积神经网络（CNN）在图像分析方面取得的巨大成功表明了将CNN的洞察力应用于点云世界的价值。点云本身缺乏拓扑信息，因此设计一个模型来恢复拓扑可以丰富点云的表示能力。为此，我们提出了一个名为EdgeConv的新神经网络模块，适用于点云上基于CNN的高级任务，包括分类和分割。EdgeConv作用于在网络的每一层中动态计算的图。它是可区分的，并且可以插入到现有的架构中。与在外部空间中操作或独立处理每个点的现有模块相比，EdgeConv具有几个吸引人的属性：它包含局部邻域信息;它可以堆叠应用于学习全局形状属性;在多层系统中，特征空间中的亲和力在原始嵌入中捕获潜在长距离的语义特征。我们展示了我们的模型在标准基准测试中的性能，包括ModelNet 40，ShapeNetPart和S3 DIS。
Point clouds provide a flexible geometric representation suitable for countless applications in computer graphics; they also comprise the raw output of most 3D data acquisition devices. While hand-designed features on point clouds have long been proposed in graphics and vision, however, the recent overwhelming success of convolutional neural networks (CNNs) for image analysis suggests the value of adapting insight from CNN to the point cloud world. Point clouds inherently lack topological information, so designing a model to recover topology can enrich the representation power of point clouds. To this end, we propose a new neural network module dubbed EdgeConv suitable for CNN-based high-level tasks on point clouds, including classification and segmentation. EdgeConv acts on graphs dynamically computed in each layer of the network. It is differentiable and can be plugged into existing architectures. Compared to existing modules operating in extrinsic space or treating each point independently, EdgeConv has several appealing properties: It incorporates local neighborhood information; it can be stacked applied to learn global shape properties; and in multi-layer systems affinity in feature space captures semantic characteristics over potentially long distances in the original embedding. We show the performance of our model on standard benchmarks, including ModelNet40, ShapeNetPart, and S3DIS.