Deep Learning for 3D Point Clouds: A Survey (2019-12-27T00:00:00.000000Z)

TL;DR

This paper presents a comprehensive review of recent progress in deep learning methods for point clouds, covering three major tasks, including 3D shape classification, 3D object detection and tracking, and 3D point cloud segmentation.

Abstract

Point cloud learning has lately attracted increasing attention due to its wide applications in many areas, such as computer vision, autonomous driving, and robotics. As a dominating technique in AI, deep learning has been successfully used to solve various 2D vision problems. However, deep learning on point clouds is still in its infancy due to the unique challenges faced by the processing of point clouds with deep neural networks. Recently, deep learning on point clouds has become even thriving, with numerous methods being proposed to address different problems in this area. To stimulate future research, this paper presents a comprehensive review of recent progress in deep learning methods for point clouds. It covers three major tasks, including 3D shape classification, 3D object detection and tracking, and 3D point cloud segmentation. It also presents comparative results on several publicly available datasets, together with insightful observations and inspiring future research directions.

Authors

Li Liu

3 papers

Bennamoun

6 papers

Yulan Guo

1 papers

TL;DR

Abstract

Authors

References276 items

Stereo Matching Using Multi-Level Cost Volume and Multi-Scale Feature Constancy

Semantic Labeling and Instance Segmentation of 3D Point Clouds Using Patch Context Analysis and Multiscale Processing

3D Instance Embedding Learning With a Structure-Aware Loss Function for Point Cloud Segmentation

View-GCN: View-Based Graph Convolutional Network for 3D Shape Analysis

Structure Aware Single-Stage 3D Object Detection From Point Cloud

P2B: Point-to-Box Network for 3D Object Tracking in Point Clouds

Instance Segmentation of LiDAR Point Clouds

Weakly Supervised Semantic Point Cloud Segmentation: Towards 10× Fewer Labels

PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation

Robust Long-Term Tracking via Instance-Specific Proposals

DALES: A Large-scale Aerial LiDAR Data Set for Semantic Segmentation

3D-MPA: Multi-Proposal Aggregation for 3D Semantic Instance Segmentation

Multi-Path Region Mining for Weakly Supervised 3D Semantic Segmentation on Point Clouds

Toronto-3D: A Large-scale Mobile LiDAR Dataset for Semantic Segmentation of Urban Roadways

OccuSeg: Occupancy-Aware 3D Instance Segmentation

Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud

Global Context Reasoning for Semantic Segmentation of 3D Point Clouds

PointASNL: Robust Point Clouds Processing Using Nonlocal Neural Networks With Adaptive Sampling

PointAugment: An Auto-Augmentation Framework for Point Cloud Classification

3DSSD: Point-Based 3D Single Stage Object Detector

Learning to Group: A Bottom-Up Framework for 3D Part Discovery in Unseen Categories

ImVoteNet: Boosting 3D Object Detection in Point Clouds With Image Votes

PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection

Object as Hotspots: An Anchor-Free 3D Object Detection Approach via Firing of Hotspots

JSNet: Joint Instance and Semantic Segmentation of 3D Point Clouds

LatticeNet: Fast Point Cloud Segmentation Using Permutohedral Lattices

Scalability in Perception for Autonomous Driving: Waymo Open Dataset

Grid-GCN for Fast and Scalable Point Cloud Learning

FlowNet3D++: Geometric Losses For Deep Scene Flow Estimation

Just Go With the Flow: Self-Supervised Scene Flow Estimation

Point Cloud Instance Segmentation using Probabilistic Embeddings

Relation Graph Network for 3D Object Detection in Point Clouds

Recent Advances in 3D Object Detection in the Era of Deep Neural Networks: A Survey

PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement

RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds

PointPainting: Sequential Fusion for 3D Object Detection

PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud Representation

RangeNet ++: Fast and Accurate LiDAR Semantic Segmentation

PointRNN: Point Recurrent Neural Network for Moving Point Cloud Processing

SRINet: Learning Strictly Rotation-Invariant Representations for Point Cloud Classification and Segmentation

Patch Refinement - Localized 3D Object Detection

MeteorNet: Deep Learning on Dynamic 3D Point Cloud Sequences

Unsupervised Multi-Task Feature Learning on Point Clouds

Learning Relationships for Multi-View 3D Object Recognition

Dynamic Points Agglomeration for Hierarchical Point Sets Learning

Multi-View PointNet for 3D Scene Understanding

Hierarchical Point-Edge Interaction Network for Point Cloud Semantic Segmentation

A*3D Dataset: Towards Autonomous Driving in Challenging Environments

DensePoint: Learning Densely Contextual Representation for Efficient Point Cloud Processing

Pooling Scores of Neighboring Points for Improved 3D Point Cloud Segmentation

Efficient Learning on Point Clouds With Basis Point Sets

A Review of Point Cloud Semantic Segmentation

ShellNet: Efficient Point Cloud Convolutional Neural Networks Using Concentric Shells Statistics

Rotation Invariant Convolutions for 3D Point Clouds Deep Learning

Interpolated Convolutional Networks for 3D Point Cloud Understanding

Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data

IoU Loss for 2D/3D Object Detection

Fast Point R-CNN

A Unified Point-Based Framework for 3D Segmentation

Dilated Point Convolutions: On the Receptive Field Size of Point Convolutions on 3D Point Clouds

Dilated Point Convolutions: On the Receptive Field of Point Convolutions

DAR-Net: Dynamic Aggregation Network for Semantic Scene Segmentation

PointAtrousGraph: Deep Hierarchical Encoder-Decoder with Point Atrous Convolution for Unorganized 3D Points

PointAtrousGraph: Deep Hierarchical Encoder-Decoder with Atrous Convolution for Point Clouds

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

Point Cloud Processing via Recurrent Set Encoding

From Points to Parts: 3D Object Detection From Point Cloud With Part-Aware and Part-Aggregation Network

Point-Voxel CNN for Efficient 3D Deep Learning

Justlookup: One Millisecond Deep Feature Extraction for Point Clouds By Lookup Tables

Effective Rotation-Invariant Point CNN with Spherical Harmonics Kernels

3D Instance Segmentation via Multi-Task Metric Learning

Image-Based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era

PyramNet: Point Cloud Pyramid Attention Network and Graph Embedding Module for Classification and Segmentation

Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds

Dominant Set Clustering and Pooling for Multi-View 3D Object Recognition

Spherical Fractal Convolutional Neural Networks for Point Cloud Recognition