LiDAR 3D Object Detection
Papers
Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks
VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection
- intro: Apple Inc
- arxiv: https://arxiv.org/abs/1711.06396
Complex-YOLO: Real-time 3D Object Detection on Point Clouds
- intro: Valeo Schalter und Sensoren GmbH & Ilmenau University of Technology
- arxiv: https://arxiv.org/abs/1803.06199
Focal Loss in 3D Object Detection
- intro: IEEE RA-L 2019
- project page: https://sites.google.com/view/fl3d
- arxiv: https://arxiv.org/abs/1809.06065
- github: https://github.com/pyun-ram/FL3D
PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud
- intro: CVPR 2019
- arxiv: https://arxiv.org/abs/1812.04244
- github(official): https://github.com/sshaoshuai/PV-RCNN
- github(official): https://github.com/sshaoshuai/PointRCNN
3D Object Detection Using Scale Invariant and Feature Reweighting Networks
- intro: AAAI 2019
- arxiv: https://arxiv.org/abs/1901.02237
3D Backbone Network for 3D Object Detection
https://arxiv.org/abs/1901.08373
Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds
https://arxiv.org/abs/1904.07537
Point-Voxel CNN for Efficient 3D Deep Learning
- intro: NeurIPS 2019 Spotlight
- project page: https://hanlab.mit.edu/projects/pvcnn/
- arxiv: https://arxiv.org/abs/1907.03739
- github: https://github.com/mit-han-lab/pvcnn
IoU Loss for 2D/3D Object Detection
- intro: 3d vision 2019
- arxiv: https://arxiv.org/abs/1908.03851
Deep Hough Voting for 3D Object Detection in Point Clouds
- intro: ICCV 2019
- intro: Facebook AI Research & Stanford University
- keywords: VoteNet
- arxiv: https://arxiv.org/abs/1904.09664
- github: https://github.com/facebookresearch/votenet
Fast Point R-CNN
- intro: ICCV 2019
- intro: CUHK & Tencent YouTu Lab
- arxiv: https://arxiv.org/abs/1908.02990
Interpolated Convolutional Networks for 3D Point Cloud Understanding
- intro: ICCV 2019
- arxiv: https://arxiv.org/abs/1908.04512
PointPillars: Fast Encoders for Object Detection from Point Clouds
- intro: nuTonomy: an APTIV company
- keywords: a single stage
- arxiv: http://http://arxiv.org/abs/1812.05784
- github(official): https://github.com/nutonomy/second.pytorch
LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving
- intro: CVPR 2019
- intro: Uber Advanced Technologies Group
- arxiv: https://arxiv.org/abs/1903.08701
Sensor Fusion for Joint 3D Object Detection and Semantic Segmentation
- intro: CVPR Workshop on Autonomous Driving 2019
- keywords: LaserNet++
- arxiv: https://arxiv.org/abs/1904.11466
Part-A^2 Net: 3D Part-Aware and Aggregation Neural Network for Object Detection from Point Cloud
From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network
- intro: TPAMI 2020
- arxiv: https://arxiv.org/abs/1907.03670
- github(official): https://github.com/sshaoshuai/PartA2-Net
- github(official): https://github.com/open-mmlab/OpenPCDet
Class-balanced Grouping and Sampling for Point Cloud 3D Object Detection
- intro: CVPR 2019
- intro: winner of nuScenes 3D Object Detection challenge in WAD
- arxiv: https://arxiv.org/abs/1908.09492
- github: https://github.com/ZhengWG/Class-balanced-Grouping-and-Sampling-for-Point-Cloud-3D-Object-Detection
End-to-End Multi-View Fusion for 3D Object Detection in LiDAR Point Clouds
- intro: CoRL 2019
- intro: Waymo LLC & Google Brain
- keywords: dynamic voxelization
- arxiv: https://arxiv.org/abs/1910.06528
SampleNet: Differentiable Point Cloud Sampling
- intro: CVPR 2020 oral
- intro: Tel Aviv University
- arxiv: https://arxiv.org/abs/1912.03663
- github: https://github.com/itailang/SampleNet
TANet: Robust 3D Object Detection from Point Clouds with Triple Attention
- intro: AAAI 2020 oral
- intro: Huazhong University of Science and Technology & Chinese Academy of Sciences
- arxiv: https://arxiv.org/abs/1912.05163
- github: https://github.com/happinesslz/TANet
PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection
- intro: CVPR 2020
- arxiv: https://arxiv.org/abs/1912.13192
- github(official): https://github.com/open-mmlab/OpenPCDet
Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud
- intro: CVPR 2020
- intro: Carnegie Mellon University
- arxiv: https://arxiv.org/abs/2003.01251
- github: https://github.com/WeijingShi/Point-GNN
PV-RCNN: The Top-Performing LiDAR-only Solutions for 3D Detection / 3D Tracking / Domain Adaptation of Waymo Open Dataset Challenges
Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations
- intro: ECCV 2020 Workshop on Perception for Autonomous Driving
- intro: University of Waterloo
- arxiv: https://arxiv.org/abs/2008.08766
- github: https://github.com/AutoVision-cloud/Deformable-PV-RCNN
PV-RCNN++: Semantical Point-Voxel Feature Interaction for 3D Object Detection
https://arxiv.org/abs/2208.13414
3DSSD: Point-based 3D Single Stage Object Detector
- intro: CVPR 2020 Oral
- arxiv: https://arxiv.org/abs/2002.10187
- github: https://github.com/Jia-Research-Lab/3DSSD
HVNet: Hybrid Voxel Network for LiDAR Based 3D Object Detection
- intro: CVPR 2020
- intro: DEEPROUTE.AI
- arxiv: https://arxiv.org/abs/2003.00186
- paper: http://openaccess.thecvf.com/content_CVPR_2020/papers/Ye_HVNet_Hybrid_Voxel_Network_for_LiDAR_Based_3D_Object_Detection_CVPR_2020_paper.pdf
SSN: Shape Signature Networks for Multi-class Object Detection from Point Clouds
- intro: ECCV 2020
- intro: CUHK & SenseTime Research & Hong Kong Baptist University
- arxiv: https://arxiv.org/abs/2004.02774
- github(mmdetection3d): https://github.com/xinge008/SSN
Range Conditioned Dilated Convolutions for Scale Invariant 3D Object Detection
- intro: Google Research & Waymo LLC
- keywords: Range Conditioned Dilation (RCD)
- arxiv: https://arxiv.org/abs/2005.09927
Deep Learning for LiDAR Point Clouds in Autonomous Driving: A Review
- intro: University of Waterloo & Sun Yat-Sen University & Xilinx Technology & Ryerson University
- arxiv: https://arxiv.org/abs/2005.09830
Structure Aware Single-stage 3D Object Detection from Point Cloud
- intro: CVPR 2020
- intro: The Hong Kong Polytechnic University & DAMO Academy, Alibaba Group
- intro: SA-SSD
- paper: http://openaccess.thecvf.com/content_CVPR_2020/papers/He_Structure_Aware_Single-Stage_3D_Object_Detection_From_Point_Cloud_CVPR_2020_paper.pdf
- paper: https://www4.comp.polyu.edu.hk/~cslzhang/paper/SA-SSD.pdf
- github: https://github.com/skyhehe123/SA-SSD
Associate-3Ddet: Perceptual-to-Conceptual Association for 3D Point Cloud Object Detection
- intro: CVPR 2020
- intro: Fudan University & Baidu Inc. & University of Science and Technology of China
- arxiv: https://arxiv.org/abs/2006.04356
SVGA-Net: Sparse Voxel-Graph Attention Network for 3D Object Detection from Point Clouds
https://arxiv.org/abs/2006.04043
Stereo RGB and Deeper LIDAR Based Network for 3D Object Detection
https://arxiv.org/abs/2006.05187
Generative Sparse Detection Networks for 3D Single-shot Object Detection
- intro: Stanford University & NVIDIA
- arxiv: https://arxiv.org/abs/2006.12356
Local Grid Rendering Networks for 3D Object Detection in Point Clouds
https://arxiv.org/abs/2007.02099
InfoFocus: 3D Object Detection for Autonomous Driving with Dynamic Information Modeling
- intro: University of Maryland & Salesforce Research
- arxiv: https://arxiv.org/abs/2007.08556
EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection
- intro: ECCV 2020
- intro: Huazhong University of Science and Technology
- arxiv: https://arxiv.org/abs/2007.08856
- github: https://github.com/happinesslz/EPNet
Pillar-based Object Detection for Autonomous Driving
- intro: ECCV 2020
- intro: MIT & Google
- arxiv: https://arxiv.org/abs/2007.10323
- github(TensorFlow): https://github.com/WangYueFt/pillar-od
Weakly Supervised 3D Object Detection from Lidar Point Cloud
- intro: ECCV 2020
- intro: Beijing Institute of Technology & ETH Zurich & Inception Institute of Artificial Intelligence
- arxiv: https://arxiv.org/abs/2007.11901
- github: https://github.com/hlesmqh/WS3D
An LSTM Approach to Temporal 3D Object Detection in LiDAR Point Clouds
- intro: ECCV 2020
- intro: Google Research
- arxiv: https://arxiv.org/abs/2007.12392
Part-Aware Data Augmentation for 3D Object Detection in Point Cloud
https://arxiv.org/abs/2007.13373
Weakly Supervised 3D Object Detection from Point Clouds
- intro: ACM MM 2020
- intro: MIT & Microsoft Research
- arxiv: https://arxiv.org/abs/2007.13970
- github: https://github.com/Zengyi-Qin/Weakly-Supervised-3D-Object-Detection
Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution
- intro: ECCV 2020
- arxiv: https://arxiv.org/abs/2007.16100
Global Context Aware Convolutions for 3D Point Cloud Understanding
https://arxiv.org/abs/2008.02986
DeepLiDARFlow: A Deep Learning Architecture For Scene Flow Estimation Using Monocular Camera and Sparse LiDAR
- intro: IROS 2020
- arxiv: https://arxiv.org/abs/2008.08136
- github: https://github.com/dfki-av/DeepLiDARFlow
PointMixup: Augmentation for Point Clouds
- intro: ECCV 2020 spotlight
- arxiv: https://arxiv.org/abs/2008.06374
Cross-Modality 3D Object Detection
- intro: WACV 2021
- arxiv: https://arxiv.org/abs/2008.10436
LC-NAS: Latency Constrained Neural Architecture Search for Point Cloud Networks
https://arxiv.org/abs/2008.10309
Multi-View Fusion of Sensor Data for Improved Perception and Prediction in Autonomous Driving
- intro: Uber Advanced Technologies Group
- arxiv: https://arxiv.org/abs/2008.11901
DV-ConvNet: Fully Convolutional Deep Learning on Point Clouds with Dynamic Voxelization and 3D Group Convolution
- intro: DESR Lab, Hong Kong University of Science and Technology
- arxiv: https://arxiv.org/abs/2009.02918
Joint Pose and Shape Estimation of Vehicles from LiDAR Data
- intro: Argo AI & Microsoft & Carnegie Mellon University
- arxiv: https://arxiv.org/abs/2009.03964
Deep Learning for 3D Point Cloud Understanding: A Survey
Multi-Frame to Single-Frame: Knowledge Distillation for 3D Object Detection
- intro: ECCV 2020 Workshop on Perception for Autonomous Driving
- intro: MIT & Google & Stanford
- arxiv: https://arxiv.org/abs/2009.11859
Torch-Points3D: A Modular Multi-Task Frameworkfor Reproducible Deep Learning on 3D Point Clouds
MLOD: Awareness of Extrinsic Perturbation in Multi-LiDAR 3D Object Detection for Autonomous Driving
- intro: The Hong Kong University of Science and Technology
- project page: https://ram-lab.com/file/site/mlod/
- arxiv: https://arxiv.org/abs/2010.11702
StrObe: Streaming Object Detection from LiDAR Packets
- intro: CoRL 2020
- intro: Uber Advanced Technologies Group & University of Toronto
- arxiv: https://arxiv.org/abs/2011.06425
MuSCLE: Multi Sweep Compression of LiDAR using Deep Entropy Models
- intro: NeurIPS 2020
- intro: Uber Advanced Technologies Group & University of Waterloo & University of Toronto
- arxiv: https://arxiv.org/abs/2011.07590
LiDAR-based Panoptic Segmentation via Dynamic Shifting Network
- intro: Nanyang Technological University & Chinese University of Hong Kong
- intro: Rank 1st place in the leaderboard of SemanticKITTI Panoptic Segmentation (accessed at 2020-11-16)
- arxiv: https://arxiv.org/abs/2011.11964
- github: https://github.com/hongfz16/DS-Net
CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloud
- intro: AAAI 2021
- intro: The Chinese University of Hong Kong
- arxiv: https://arxiv.org/abs/2012.03015
- github: https://github.com/Vegeta2020/CIA-SSD
PC-RGNN: Point Cloud Completion and Graph Neural Network for 3D Object Detection
- intro: Beihang University
- arxiv: https://arxiv.org/abs/2012.10412
Achieving Real-Time LiDAR 3D Object Detection on a Mobile Device
https://arxiv.org/abs/2012.13801
Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection
- intro: AAAI 2021
- arxiv: https://arxiv.org/abs/2012.15712
- github: https://github.com/djiajunustc/Voxel-R-CNN
RTS3D: Real-time Stereo 3D Detection from 4D Feature-Consistency Embedding Space for Autonomous Driving
Self-Attention Based Context-Aware 3D Object Detection
- intro: University of Waterloo
- arxiv: https://arxiv.org/abs/2101.02672
- github: https://github.com/AutoVision-cloud/SA-Det3D
A Simple and Efficient Multi-task Network for 3D Object Detection and Road Understanding
ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detection
- intro: CVPR 2021
- arxiv: https://arxiv.org/abs/2103.05346
- github: https://github.com/CVMI-Lab/ST3D
RangeDet:In Defense of Range View for LiDAR-based 3D Object Detection
- intro: ICCV 2021
- arxiv: https://arxiv.org/abs/2103.10039
- github: https://github.com/TuSimple/RangeDet
Stereo CenterNet based 3D Object Detection for Autonomous Driving
https://arxiv.org/abs/2103.11071
LiDAR R-CNN: An Efficient and Universal 3D Object Detector
- intro: CVPR 2021
- intro: TuSimple
- arxiv: https://arxiv.org/abs/2103.15297
- github: https://github.com/tusimple/LiDAR_RCNN
HVPR: Hybrid Voxel-Point Representation for Single-stage 3D Object Detection
- intro: CVPR 2021 https://arxiv.org/abs/2104.00902
Group-Free 3D Object Detection via Transformers
- intro: University of Science and Technology of China & Microsoft Research Asia
- arxiv: https://arxiv.org/abs/2104.00678
- github: https://github.com/zeliu98/Group-Free-3D
SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud
- intro: CVPR 2021
- arxiv: https://arxiv.org/abs/2104.09804
- github: https://github.com/Vegeta2020/SE-SSD
BEVDetNet: Bird’s Eye View LiDAR Point Cloud based Real-time 3D Object Detection for Autonomous Driving
https://arxiv.org/abs/2104.10780
Investigating Attention Mechanism in 3D Point Cloud Object Detection
- intro: Australian National University & Data61-CSIRO, Australia & University of Technology Sydney & Nanyang Technological University
- arxiv: https://arxiv.org/abs/2108.00620
- github: https://github.com/ShiQiu0419/attentions_in_3D_detection
Pyramid R-CNN: Towards Better Performance and Adaptability for 3D Object Detection
- intro: ICCV 2021
- intro: The Chinese University of Hong Kong & Huawei Noah’s Ark Lab & HKUST & Sun Yat-Sen University
- arxiv: https://arxiv.org/abs/2109.02499
Voxel Transformer for 3D Object Detection
- intro: ICCV 2021
- intro: The Chinese University of Hong Kong & National University of Singapore & Huawei Noah’s Ark Lab & HKUST & Sun Yat-Sen University
- arxiv: https://arxiv.org/abs/2109.02497
A Versatile Multi-View Framework for LiDAR-based 3D Object Detection with Guidance from Panoptic Segmentation
- intro: CVPR 2022
- arxiv: https://arxiv.org/abs/2203.02133
Point Density-Aware Voxels for LiDAR 3D Object Detection
- intro: CVPR 2022
- arxiv: https://arxiv.org/abs/2203.05662
- github: https://github.com/TRAILab/PDV
VISTA: Boosting 3D Object Detection via Dual Cross-VIew SpaTial Attention
- intro: CVPR 2022
- arxiv: https://arxiv.org/abs/2203.09704
- github: https://github.com/Gorilla-Lab-SCUT/VISTA
Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds
- intro: CVPR 2022
- intro: National University of Defense Technology & University of Oxford
- arxiv: https://arxiv.org/abs/2203.11139
- github: https://github.com/yifanzhang713/IA-SSD
LiDAR Distillation: Bridging the Beam-Induced Domain Gap for 3D Object Detection
- intro: Tsinghua University & State Key Lab of Intelligent Technologies and Systems & Gaussian Robotics
- arxiv: https://arxiv.org/abs/2203.14956
- github: https://github.com/weiyithu/LiDAR-Distillation
Point2Seq: Detecting 3D Objects as Sequences
- intro: CVPR 2022
- arxiv: https://arxiv.org/abs/2203.13394
- github: https://github.com/ocNflag/point2seq
OccAM’s Laser: Occlusion-based Attribution Maps for 3D Object Detectors on LiDAR Data
- intro: CVPR 2022
- arxiv: https://arxiv.org/abs/2204.06577
- github: https://github.com/dschinagl/occam
PointDistiller: Structured Knowledge Distillation Towards Efficient and Compact 3D Detection
- intro: Tsinghua University & Xi’an Jiaotong University & DIDI
- arxiv: https://arxiv.org/abs/2205.11098
- github: https://github.com/RunpeiDong/PointDistiller
PillarNet: Real-Time and High-Performance Pillar-based 3D Object Detection
- intro: Harbin Institute of Technology & Shanghai Jiao Tong University
- arxiv: https://arxiv.org/abs/2205.07403
Fully Convolutional One-Stage 3D Object Detection on LiDAR Range Images
- intro: Meituan Inc. & Zhejiang University & Northwestern Polytechnical University
- arxiv: https://arxiv.org/abs/2205.13764
Voxel Field Fusion for 3D Object Detection
- intro: CVPR 2022
- intro: The Chinese University of Hong Kong & The University of Hong Kong & MEGVII Technology & SmartMore
- arxiv: https://arxiv.org/abs/2205.15938
- github: https://github.com/dvlab-research/VFF
Unifying Voxel-based Representation with Transformer for 3D Object Detection
- intro: The Chinese University of Hong Kong & The University of Hong Kong & MEGVII Technology & SmartMore4
- arxiv: https://arxiv.org/abs/2206.00630
- github: https://github.com/dvlab-research/UVTR
LidarMultiNet: Unifying LiDAR Semantic Segmentation, 3D Object Detection, and Panoptic Segmentation in a Single Multi-task Network
- intro: TuSimple & University of Central Florida
- intro: Official 1st Place Solution for the Waymo Open Dataset Challenges 2022 - 3D Semantic Segmentation
- arxiv: https://arxiv.org/abs/2206.11428
Rethinking IoU-based Optimization for Single-stage 3D Object Detection
- intro: ECCV 2022
- intro: Zhejiang University & Alibaba Cloud Computing Ltd. & National University of Singapore
- arxiv: https://arxiv.org/abs/2207.09332
- github: https://github.com/hlsheng1/RDIoU
Embracing Single Stride 3D Object Detector with Sparse Transformer
- intro: CVPR 2022
- intro: CASIA & UIUC & CMU & THU & TuSimple
- arxiv: https://arxiv.org/abs/2112.06375
- github: https://github.com/TuSimple/SST
- zhihu: https://zhuanlan.zhihu.com/p/476056546
Fully Sparse 3D Object Detection
- intro: ECCV 2022
- intro: CASIA & TuSimple
- arxiv: https://arxiv.org/abs/2207.10035
- github: https://github.com/TuSimple/SST
Anchor-free 3D Detection
Object as Hotspots: An Anchor-Free 3D Object Detection Approach via Firing of Hotspots
- intro: Samsung Inc & Johns Hopkins University & South China University of Technology
- keywords: Object as Hotspots (OHS)
- arxiv: https://arxiv.org/abs/1912.12791
CenterNet3D: An Anchor free Object Detector for Autonomous Driving
- keywords: Non-Maximum Suppression free
- arxiv: https://arxiv.org/abs/2007.07214
- github: https://github.com/wangguojun2018/CenterNet3d
AFDet: Anchor Free One Stage 3D Object Detection
- intro: Horizon Robotics
- intro: CVPR Workshop 2020
- intro: Baseline detector for the 1st place solutions of Waymo Open Dataset Challenges 2020
- arxiv: https://arxiv.org/abs/2006.12671
Real-Time Anchor-Free Single-Stage 3D Detection with IoU-Awareness
- intro: Horizon Robotics
- keywords: AFDetV2
- arxiv: https://arxiv.org/abs/2107.14342
1st Place Solution for Waymo Open Dataset Challenge – 3D Detection and Domain Adaptation
- intro: Horizon Robotics
- arxiv: https://arxiv.org/abs/2006.15505
FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection
- intro: Samsung AI Center Moscow
- arxiv: https://arxiv.org/abs/2112.00322
- github: https://github.com/samsunglabs/fcaf3d
3D Semantic Segmentation
PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation
- intro: CVPR 2020
- arxiv: https://arxiv.org/abs/2003.14032
- github: https://github.com/edwardzhou130/PolarSeg
Cloud Transformers
- intro: Samsung AI Center Moscow & Skolkovo Institute of Science and Technology
- arxiv: https://arxiv.org/abs/2007.11679
Cylinder3D: An Effective 3D Framework for Driving-scene LiDAR Semantic Segmentation
- intro: CUHK & ShanghaiTech University & SenseTime Research
- arxiv: https://arxiv.org/abs/2008.01550
- github: https://github.com/xinge008/Cylinder3D
Projected-point-based Segmentation: A New Paradigm for LiDAR Point Cloud Segmentation
https://arxiv.org/abs/2008.03928
pseudo-LiDAR
Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving
- intro: CVPR 2019
- project page: https://mileyan.github.io/pseudo_lidar/
- arxiv: https://arxiv.org/abs/1812.07179
- gtihub(official): https://github.com/mileyan/pseudo_lidar
Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving
- intro: ICLR 2020 Poster
- openreview: https://openreview.net/forum?id=BJedHRVtPB
- github(official): https://github.com/mileyan/Pseudo_Lidar_V2
End-to-End Pseudo-LiDAR for Image-Based 3D Object Detection
- intro: CVPR 2020
- arxiv: https://arxiv.org/abs/2004.03080
- github: https://github.com/mileyan/pseudo-LiDAR_e2e
Rethinking Pseudo-LiDAR Representation
- intro: ECCV 2020
- arxiv: https://arxiv.org/abs/2008.04582
- github: https://github.com/xinzhuma/patchnet
Demystifying Pseudo-LiDAR for Monocular 3D Object Detection
- intro: University of Trento & Fondazione Bruno Kessler & Facebook
- arxiv: https://arxiv.org/abs/2012.05796
Is Pseudo-Lidar needed for Monocular 3D Object detection?
- intro: ICCV 2021
- intro: Toyota Research Institute
- arxiv: https://arxiv.org/abs/2108.06417
ProposalContrast: Unsupervised Pre-training for LiDAR-based 3D Object Detection
- intro: Beijing Institute of Technology & Baidu Research & National Engineering Laboratory of Deep Learning Technology and Application & University of Macau & University of Technology Sydney
- arxiv: https://arxiv.org/abs/2207.12654
- github: https://github.com/yinjunbo/ProposalContrast
TransPillars: Coarse-to-Fine Aggregation for Multi-Frame 3D Object Detection
- intro: Nanyang Technological University & Sensetime Research
- arxiv: https://arxiv.org/abs/2208.03141
Multi-Modal 3D Object Detection
- intro: University of Science and Technology & Harbin Institute of Technology & SenseTime Research & The Chinese University of Hong Kong & IIIS, Tsinghua University AutoAlign: Pixel-Instance Feature Aggregation for Multi-Modal 3D Object Detection
- arxiv: https://arxiv.org/abs/2201.06493
3D Detection and Tracking
Joint Monocular 3D Vehicle Detection and Tracking
- intro: ICCV 2019
- project page: https://eborboihuc.github.io/Mono-3DT/
- arxiv: https://arxiv.org/abs/1811.10742
- github(official): https://github.com/ucbdrive/3d-vehicle-tracking
Center-based 3D Object Detection and Tracking
- intro: UT Austin
- intro: 3D Object Detection and Tracking using center points in the bird-eye view.
- arxiv: https://arxiv.org/abs/2006.11275
- github: https://github.com/tianweiy/CenterPoint
3D Object Detection and Tracking Based on Streaming Data
- intro: ICRA 2020
- arxiv: https://arxiv.org/abs/2009.06169
Uncertainty-Aware Voxel based 3D Object Detection and Tracking with von-Mises Loss
- intro: University of Michigan
- arxiv: https://arxiv.org/abs/2011.02553
Joint Multi-Object Detection and Tracking with Camera-LiDAR Fusion for Autonomous Driving
- intro: IROS 2021
- arxiv: https://arxiv.org/abs/2108.04602
PTT: Point-Track-Transformer Module for 3D Single Object Tracking in Point Clouds
- intro: IROS 2021
- arxiv: https://arxiv.org/abs/2108.06455
- github: https://github.com/shanjiayao/PTT
Real-time 3D Single Object Tracking with Transformer
- intro: IEEE Transactions on Multimedia
- arxiv: https://arxiv.org/abs/2209.00860
- github: https://github.com/shanjiayao/PTT
3D MOT
AutoSelect: Automatic and Dynamic Detection Selection for 3D Multi-Object Tracking
- intro: Carnegie Mellon University
- arxiv: https://arxiv.org/abs/2012.05894
Probabilistic 3D Multi-Modal, Multi-Object Tracking for Autonomous Driving
- intro: Stanford University & Toyota Research Institute
- arxiv: https://arxiv.org/abs/2012.13755
Monocular Quasi-Dense 3D Object Tracking
https://arxiv.org/abs/2103.07351
Lite-FPN for Keypoint-based Monocular 3D Object Detection
RSN: Range Sparse Net for Efficient, Accurate LiDAR 3D Object Detection
- intro: CVPR 2021
- intro: Waymo LLC & Google
- arxiv: https://arxiv.org/abs/2106.13365
VIN: Voxel-based Implicit Network for Joint 3D Object Detection and Segmentation for Lidars
https://arxiv.org/abs/2107.02980
Geometry Uncertainty Projection Network for Monocular 3D Object Detection
- intro: ICCV 2021
- arxiv: https://arxiv.org/abs/2107.13774
Exploring Simple 3D Multi-Object Tracking for Autonomous Driving
- intro: ICCV 2021
- intro: QCraft & Johns Hopkins University
- arxiv: https://arxiv.org/abs/2108.10312
CAMO-MOT: Combined Appearance-Motion Optimization for 3D Multi-Object Tracking with Camera-LiDAR Fusion
- intro: Based on this work, we achieved 1st place on the nuScenes tracking leaderboard
- intro: Tsinghua University & Mogo Auto
- arxiv: https://arxiv.org/abs/2209.02540
Transformer
Point Transformer
- intro: Ulm University
- keywords: SortNet
- arxiv: https://arxiv.org/abs/2011.00931
Temporal-Channel Transformer for 3D Lidar-Based Video Object Detection in Autonomous Driving
https://arxiv.org/abs/2011.13628
Point Transformer
- intro: University of Oxford & The Chinese University of Hong Kong & Intel Labs
- arxiv: https://arxiv.org/abs/2012.09164
3D Object Detection with Pointformer
- intro: Tsinghua University & BNRist & Alexa AI, Amazon / Columbia University
- arxiv: https://arxiv.org/abs/2012.11409
M3DeTR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers
- intro: University of Maryland & Fudan University
- arxiv: https://arxiv.org/abs/2104.11896
Improving 3D Object Detection with Channel-wise Transformer
- intro: ICCV 2021
- intro: Zhejiang University & DAMO Academy, Alibaba Group
- arxiv: https://arxiv.org/abs/2108.10723
- github: https://github.com/hlsheng1/CT3D
TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with Transformers
- intro CVPR 2022
- intro: Hong Kong University of Science and Technology & Huawei & City University of Hong Kong
- arxiv: https://arxiv.org/abs/2203.11496
- github: https://github.com/XuyangBai/TransFusion/
Projects
OpenLidarPerceptron
- intro: OpenLidarPerceptron is an open source project for LiDAR-based 3D scene perception.
- github: https://github.com/open-mmlab/OpenLidarPerceptron
Super Fast and Accurate 3D Object Detection based on 3D LiDAR Point Clouds
- github(PyTorch): https://github.com/maudzung/SFA3D
Resources
Awesome-Automanous-3D-Detection-Methods
https://github.com/tyjiang1997/awesome-Automanous-3D-detection-methods