小执着的博客

CVPR2020 3D点云相关论文整理3D目标检测1. A Hierarchical Graph Network for 3D Object Detection on Point Clouds1.1 创新1.2 实现细节1.3 实验结果3D点云分析（分类、分割）3D点云上采样1. Cascaded Refinement Network for Point Cloud Completion1.1 创新1.2 实现细节1.3 实验结果3D目标检测1. A Hierarchical Graph Network

PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud一、论文思路二、模型实现2.1 阶段一2.2 阶段二2.3 实现细节三、实验结果代码论文一、论文思路本文提出了一个两阶段的3D detection模型PointRCNN。论文的模型分为两个阶段，第一阶段先使用PointNet++作为一个前景分割模型，...

《Frustum PointNets for 3D Object Detection from RGB-D Data》论文笔记一、论文思路二、模型介绍2.1 模型结构2.2 实现细节2.2.1 Frustum Proposal2.2.2 3D Instance Segmentation2.2.3 Amodal 3D Box Estimation2.2.4 模型结构2.2.5 损失函数三、实验结...

Deep Hough Voting for 3D Object Detection in Point Clouds一、论文贡献二、模型实现2.1 霍夫投票2.2 VoteNet2.2.1 总体架构2.2.2 Point cloud feature learning2.2.3 Hough voting with deep networks2.2.4 Object Proposal and Clas...

PIXOR: Real-time 3D Object Detection from Point Clouds 阅读笔记文章标题PIXOR: Real-time 3D Object Detection from Point Clouds作者Bin Yang, Wenjie Luo, Raquel Urtasun来源IEEE Conference on Computer Vision an...