PC-RGNN: Point Cloud Completion and Graph Neural Network for 3D Object Detection
This work provides a strong specific gain in 3D object detection performance for autonomous driving by improving handling of sparse and partial point clouds.
The paper addresses the challenge of sparse and partial point clouds in LiDAR-based 3D object detection for autonomous driving. They propose PC-RGNN, a two-stage approach that recovers high-quality point cloud proposals and uses a graph neural network to strengthen encoded features, achieving remarkable improvements on the KITTI benchmark.
LiDAR-based 3D object detection is an important task for autonomous driving and current approaches suffer from sparse and partial point clouds of distant and occluded objects. In this paper, we propose a novel two-stage approach, namely PC-RGNN, dealing with such challenges by two specific solutions. On the one hand, we introduce a point cloud completion module to recover high-quality proposals of dense points and entire views with original structures preserved. On the other hand, a graph neural network module is designed, which comprehensively captures relations among points through a local-global attention mechanism as well as multi-scale graph based context aggregation, substantially strengthening encoded features. Extensive experiments on the KITTI benchmark show that the proposed approach outperforms the previous state-of-the-art baselines by remarkable margins, highlighting its effectiveness.