3D object detection is receiving increasing attention from both industry and
academia thanks to its wide applications in various fields. In this paper, we
propose Point-Voxel Region-based Convolution Neural Networks (PV-RCNNs) for 3D
object detection on point clouds. First, we propose a novel 3D detector,
PV-RCNN, which boosts the 3D detection performance by deeply integrating the
feature learning of both point-based set abstraction and voxel-based sparse
convolution through two novel steps, i.e., the voxel-to-keypoint scene encoding
and the keypoint-to-grid RoI feature abstraction. Second, we propose an
advanced framework, PV-RCNN++, for more efficient and accurate 3D object
detection. It consists of two major improvements: sectorized proposal-centric
sampling for efficiently producing more representative keypoints, and
VectorPool aggregation for better aggregating local point features with much
less resource consumption. With these two strategies, our PV-RCNN++ is about
$3\times$ faster than PV-RCNN, while also achieving better performance. The
experiments demonstrate that our proposed PV-RCNN++ framework achieves
state-of-the-art 3D detection performance on the large-scale and
highly-competitive Waymo Open Dataset with 10 FPS inference speed on the
detection range of 150m * 150m.

本文提出了一种针对点云的三维目标检测新方法 PV-RCNN，通过点和体素两个方面来深度整合特征学习，采用体素到关键点场景编码和关键点到网格 RoI 特征抽象两个创新步骤，更高效准确地实现三维目标检测。PV-RCNN++ 框架通过部分区域划分建议 - 中心采样和 VectorPool 池化等两种策略，在更低资源消耗下实现了更优秀的性能，相比 PV-RCNN 在保持准确率的同时提高了 3 倍速度，并在大规模 Waymo 开放数据集上取得了最先进的三维检测性能。