自动驾驶点云处理文章(转自github)

  • Post author:
  • Post category:其他





Lidar Point clound processing for Autonomous Driving

A list of references on lidar point cloud processing for autonomous driving




LiDAR Pointcloud Clustering/Semantic Segmentation


Tasks

: Road/Ground extraction, plane extraction, Semantic segmentation, open set instance segmentation, Clustering

  • Fast Segmentation of 3D Point Clouds: A Paradigm on LiDAR Data for Autonomous Vehicle Applications [

    git

    ]
  • Time-series LIDAR Data Superimposition for Autonomous Driving [

    pdf

    ]
  • Fast segmentation of 3D point clouds for ground vehicles [

    ieee

    ]
  • An Improved RANSAC for 3D Point Cloud Plane Segmentation Based on Normal Distribution Transformation Cells
  • Segmentation of Dynamic Objects from Laser Data [

    pdf

    ]
  • A Fast Ground Segmentation Method for 3D Point Cloud [

    pdf

    ]
  • Ground Estimation and Point Cloud Segmentation using SpatioTemporal Conditional Random Field [

    pdf

    ]
  • Real-Time Road Segmentation Using LiDAR Data Processing on an FPGA [

    pdf

    ]
  • Efficient Online Segmentation for Sparse 3D Laser Scans [

    pdf

    ], [

    git

    ]
  • CNN for Very Fast Ground Segmentation in Velodyne LiDAR Data [

    pdf

    ]
  • A Comparative Study of Segmentation and Classification Methods for 3D Point Clouds [

    pdf

    ]
  • Fast Multi-pass 3D Point Segmentation Based on a Structured Mesh Graph for Ground Vehicles

    pdf


    video
  • Circular Convolutional Neural Networks for Panoramic Images and Laser Data

    pdf
  • Efficient Convolutions for Real-Time Semantic Segmentation of 3D Point Clouds [

    pdf

    ]
  • Identifying Unknown Instances for Autonomous Driving/Open-set instance segmentation algorithm

    CoRL 2019

    [

    pdf

    ]
  • RIU-Net: Embarrassingly simple semantic segmentation of3D LiDAR point cloud. [

    pdf

    ,

    LU-net

    ]
  • SalsaNet: Fast Road and Vehicle Segmentation in LiDAR Point Clouds for Autonomous Driving [

    pdf

    ]




Continous domain DNNs

  • Deep Parametric Continuous Convolutional Neural Networks CVPR 2018 [

    pdf

    ]




Pointcloud Density

  • DBSCAN : A density-based algorithm for discovering clusters in large spatial databases with noise (1996) [

    pdf

    ]
  • Hierarchical Density Estimates for Data Clustering, Visualization, and Outlier Detection

    pdf
  • Building Maps for Autonomous Navigation Using Sparse Visual SLAM Features [

    pdf

    ]
  • STD: Sparse-to-Dense 3D Object Detector for Point Cloud

    pdf
  • Fast semantic segmentation of 3d point clounds with strongly varying density [

    pdf

    ]
  • The Perfect Match: 3D Point Cloud Matching with Smoothed Densities [

    pdf

    ,

    code

    ]




Registration and Localization

  • Point Clouds Registration with Probabilistic Data Association [

    git

    ]
  • Robust LIDAR Localization using Multiresolution Gaussian Mixture Maps for Autonomous Driving [

    pdf

    ], [

    Thesis

    ]
  • Automatic Merging of Lidar Point-Clouds Using Data from Low-Cost GPS/IMU Systems [

    pdf

    ]
  • Fast and Robust 3D Feature Extraction from Sparse Point Clouds [

    pdf

    ]
  • 3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration [

    pdf

    ]
  • Incremental Segment-Based Localization in 3D Point Clouds [

    pdf

    ]




Feature Extraction

  • Fast Feature Detection and Stochastic Parameter Estimation of Road Shape using Multiple LIDAR [

    pdf

    ]
  • Finding Planes in LiDAR Point Clouds for Real-Time Registration [

    pdf

    ]
  • Online detection of planes in 2D lidar [

    pdf

    ]
  • A Fast RANSAC–Based Registration Algorithm for Accurate Localization in Unknown Environments using LIDAR Measurements [

    pdf

    ]
  • Hierarchical Plane Extraction (HPE): An Efficient Method For Extraction Of Planes From Large Pointcloud Datasets [

    pdf

    ]
  • A Fast and Accurate Plane Detection Algorithm for Large Noisy Point Clouds Using Filtered Normals and Voxel Growing [

    pdf

    ]
  • SPLATNet: Sparse Lattice Networks for Point Cloud Processing CVPR 2018 [

    pdf

    ,

    code

    ]




Object detection and Tracking

  • Learning a Real-Time 3D Point Cloud Obstacle Discriminator via Bootstrapping

    pdf
  • Terrain-Adaptive Obstacle Detection [

    pdf

    ]
  • 3D Object Detection from Roadside Data Using Laser Scanners [

    pdf

    ]
  • 3D Multiobject Tracking for Autonomous Driving : Masters thesis A S Abdul Rahman
  • Motion-based Detection and Tracking in 3D LiDAR Scans [

    pdf

    ]
  • Lidar-histogram for fast road and obstacle detection [

    pdf

    ]
  • End-to-end Learning of Multi-sensor 3D Tracking by Detection

    pdf
  • Leveraging Heteroscedastic Aleatoric Uncertainties for Robust Real-Time LiDAR 3D Object Detection

    pdf
  • Deep tracking in the wild: End-to-end tracking using recurrent neural networks [

    pdf

    ]
  • Leveraging Heteroscedastic Aleatoric Uncertainties for Robust Real-Time LiDAR 3D Object Detection [

    pdf

    ], [

    video

    ]
  • VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection CVPR 2018 [

    pdf

    ,

    code

    ]
  • PIXOR: Real-time 3D Object Detection from Point Clouds CVPR 2018 [

    pdf

    ]
  • Deep Multi-modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challenges [

    pdf

    ]
  • Low resolution lidar-based multi-object tracking for driving applications [

    pdf

    ]
  • Patch Refinement — Localized 3D Object Detection [

    pdf

    ]




Classification/Supervised Learning

  • PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation [

    link

    ,

    link2

    ]
  • SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloud

    pdf
  • Improving LiDAR Point Cloud Classification using Intensities and Multiple Echoes [

    pdf

    ]
  • DepthCN: Vehicle Detection Using 3D-LIDAR and ConvNet [

    pdf

    ]
  • 3D Object Localisation with Convolutional Neural Networks [

    Thesis

    ]
  • SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud [

    pdf

    ]
  • PointSeg: Real-Time Semantic Segmentation Based on 3D LiDAR Point Cloud [

    pdf

    ]
  • Fast LIDAR-based Road Detection Using Fully Convolutional Neural Networks [

    pdf

    ]
  • ChipNet: Real-Time LiDAR Processing for Drivable Region Segmentation on an FPGA [

    pdf

    ]




Maps / Grids / HD Maps / Occupancy grids/ Prior Maps

  • Hierarchies of Octrees for Efficient 3D Mapping

    pdf
  • Adaptive Resolution Grid Mapping using Nd-Tree [

    ieee

    ], [

    pdf

    ], [

    video

    ]
  • LIDAR-Data Accumulation Strategy To Generate High Definition Maps For Autonomous Vehicles [

    link

    ]
  • Long-term 3D map maintenance in dynamic environments [

    video

    ]
  • Detection and Tracking of Moving Objects Using 2.5D Motion Grids [

    pdf

    ]
  • 3D Lidar-based Static and Moving Obstacle Detection in Driving Environments: an approach based on voxels and multi-region ground planes [

    pdf

    ]
  • Spatio–Temporal Hilbert Maps for Continuous Occupancy Representation in Dynamic Environments [

    pdf

    ]
  • Dynamic Occupancy Grid Prediction for Urban Autonomous Driving: A Deep Learning Approach with Fully Automatic Labeling [

    pdf

    ]
  • Fast 3-D Urban Object Detection on Streaming Point Clouds [

    pdf

    ]
  • Mobile Laser Scanned Point-Clouds for Road Object Detection and Extraction: A Review [

    pdf

    ]
  • Efficient Continuous-time SLAM for 3D Lidar-based Online Mapping [

    pdf

    ]
  • DeLS-3D: Deep Localization and Segmentation with a 3D Semantic Map [

    pdf

    ],[

    video

    ]
  • Recurrent-OctoMap: Learning State-based Map Refinement for Long-Term Semantic Mapping with 3D-Lidar Data [

    pdf

    ]
  • HDNET: Exploiting HD Maps for 3D Object Detection [

    pdf

    ]




End-To-End Learning

  • Monocular Fisheye Camera Depth Estimation Using Semi-supervised Sparse Velodyne Data [

    pdf

    ]
  • Fast and Furious: Real Time End-to-End 3D Detection, Tracking and Motion Forecasting with a Single Convolutional Net [

    pdf

    ]




Lidar Datasets and Simulators

  • nuScenes : public large-scale dataset for autonomous driving [

    dataset

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

    pdf

    ]
  • Ford Campus Vision and Lidar Data Set [

    pdf

    ,

    dataset

    ]
  • Oxford RobotCar dataset

    dataset

    1 Year, 1000km: The Oxford RobotCar Dataset

    pdf
  • Udacity based simulator [

    link

    ,

    git

    ]
  • Tutorial on Gazebo to simulate raycasting from Velodyne lidar [

    link

    ]
  • Udacity Driving Dataset [

    link

    ]
  • Virtual KITTI [

    link

    ]
  • LiDAR-Video Driving Dataset: Learning Driving Policies Effectively [

    pdf

    ]
  • KAIST Complex Urban Data Set Dataset [

    dataset

    ]
  • Semantic KITTI [

    dataset

    ]
  • A*3D: An Autonomous Driving Dataset in Challeging Environments [

    dataset

    ], [

    video

    ]




Spatio-Temporal, Movement, Flow estimation in Pointclouds

  • Rigid Scene Flow for 3D LiDAR Scans IROS 2016 [

    pdf

    ]
  • Deep Lidar CNN to Understand the Dynamics of Moving Vehicles [

    pdf

    ]
  • Learning motion field of LiDAR point cloud with convolutional networks [

    link

    ]
  • Hallucinating Dense Optical Flow from Sparse Lidar for Autonomous Vehicles [

    pdf


    video

    ]
  • FlowNet3D: Learning Scene Flow in 3D Point Clouds CVPR 2019 [

    pdf

    ,

    code

    ]
  • LiDAR-Flow: Dense Scene Flow Estimation from Sparse LiDAR and Stereo Images [

    pdf

    ]
  • 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks CVPR 2019 [

    pdf

    ,

    code

    ]
  • MeteorNet: Deep Learning on Dynamic 3D Point Cloud Sequences, ICCV 2019 [

    pdf

    ]




Advanced Topics/Other applications


TAsks

: Upsampling, Domain adaptation Sim2Real, NAS, SSL

  • PU-GAN: a Point Cloud Upsampling Adversarial Network ICCV 2019 [

    pdf

    ,

    code

    ]
  • Neural Architecture Search for Object Detection in Point Cloud [

    blog

    ], [

    AutoDeepLabNAS paper

    ]
  • Self-Supervised Deep Learning on Point Clouds by Reconstructing Space NeurIPS 2019 [

    pdf

    ]
  • Domain Adaptation for Vehicle Detection from Bird’s Eye View LiDAR Point Cloud Data ICCVW 2019

    pdf
  • Weighted Point Cloud Augmentation for Neural Network Training Data Class-Imbalance [

    pdf

    ]




Large-scale pointcloud Algorithms (vs scan based)

  • Datasets :

    • Semantic 3D

      dataset
    • Paris-Lille-3D: A Point Cloud Dataset for Urban Scene Segmentation and Classification [

      pdf


      dataset

      ]
    • SynthCity: A large-scale synthetic point cloud 2019 [

      dataset

      ,

      pdf

      ]
    • HD Map Dataset & Localization Dataset NAVER Labs : [

      link

      ]
  • Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs CVPR2018 [

    pdf

    ]
  • PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space NeurIPS 2017 [

    pdf

    ,

    code

    ],

    semantic seg code
  • ConvPoint: continuous convolutions for cloud processing Eurographics 3DOR, 2019

    pdf

    ,

    code
  • Classification of Point Cloud for Road Scene Understanding with Multiscale Voxel Deep Network

    Slides
  • Semantic Segmentation of 3D point Clouds Loic Landireu [

    Slides

    ]
  • Point Cloud Oversegmentation with Graph-Structured Deep Metric Learning [

    CVPR Workshop 2019

    ],

    video




Tools/SW/Packages