Large-scale 3D point cloud representations via graph inception networks with applications to autonomous driving

06/26/2019
by   Siheng Chen, et al.
0

We present a novel graph-neural-network-based system to effectively represent large-scale 3D point clouds with the applications to autonomous driving. Many previous works studied the representations of 3D point clouds based on two approaches, voxelization, which causes discretization errors and learning, which is hard to capture huge variations in large-scale scenarios. In this work, we combine voxelization and learning: we discretize the 3D space into voxels and propose novel graph inception networks to represent 3D points in each voxel. This combination makes the system avoid discretization errors and work for large-scale scenarios. The entire system for large-scale 3D point clouds acts like the blocked discrete cosine transform for 2D images; we thus call it the point cloud neural transform (PCT). We further apply the proposed PCT to represent real-time LiDAR sweeps produced by self-driving cars and the PCT with graph inception networks significantly outperforms its competitors.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
12/29/2018

Feature Preserving and Uniformity-controllable Point Cloud Simplification on Graph

With the development of 3D sensing technologies, point clouds have attra...
research
03/31/2018

A LiDAR Point Cloud Generator: from a Virtual World to Autonomous Driving

3D LiDAR scanners are playing an increasingly important role in autonomo...
research
09/13/2022

SVNet: Where SO(3) Equivariance Meets Binarization on Point Cloud Representation

Efficiency and robustness are increasingly needed for applications on 3D...
research
11/21/2020

MoNet: Motion-based Point Cloud Prediction Network

Predicting the future can significantly improve the safety of intelligen...
research
11/22/2020

Nudge Attacks on Point-Cloud DNNs

The wide adaption of 3D point-cloud data in safety-critical applications...
research
06/27/2020

Learning to Detect 3D Objects from Point Clouds in Real Time

In this work, we address the problem of 3D object detection from point c...
research
03/16/2018

The ApolloScape Dataset for Autonomous Driving

Scene parsing aims to assign a class (semantic) label for each pixel in ...

Please sign up or login with your details

Forgot password? Click here to reset