Robust-PointPillars 3-D Target Detection Based on Global Attention Machanism
A Robust-PointPillars 3-D target detection method based on global attention machanism is proposed,which improves the accuracy and robustness of target detection in the application of in-telligent driving.PointPillars and other neural networks have the potential to enable 3-D target de-tection by using pillars to represent the point cloud.This paper firstly introduces the dual attention module of space and channel for enhancing the feature of point cloud with learning value,which solves the lack of learning mechanism and feature extraction for Pointpillars.Then squeeze-and-ex-citation networks(SENet)module is introduced to further improve the learning and understanding ability of PointPillars to feature information.Finally,the disturbed or missing sensor signals are re-strained and the global attention algorithm is used to improve the robustness.According to the tar-get detection results based on KITTI data set,the algorithm proposed in this paper has excellent target detection accuracy and robustness.
NETL
NSTL
万方数据
