3D - Point Cloud

Paper list and Datasets about Point Cloud. Datasets can be found in Datasets.md.

Survey papers

• 3D Object Detection for Autonomous Driving: A Comprehensive Survey [IJCV 2023; Github]
• Unsupervised Point Cloud Representation Learning with Deep Neural Networks: A Survey [TPAMI 2023; Github]
• A Survey of Label-Efficient Deep Learning for 3D Point Clouds [arXiv 2023; Github]
• Self-Supervised Learning for Point Clouds Data: A Survey [arXiv 2023]
• Self-supervised Learning for Pre-Training 3D Point Clouds: A Survey [arXiv 2023]
• Radar-Camera Fusion for Object Detection and Semantic Segmentation in Autonomous Driving: A Comprehensive Review [arXiv 2023; Project]
• Perception Datasets for Anomaly Detection in Autonomous Driving: A Survey [arXiv 2023]
• Delving into the Devils of Bird's-eye-view Perception: A Review, Evaluation and Recipe [arXiv 2022; Github]
• 3D Vision with Transformers: A Survey [arXiv 2022; Github]
• Vision-Centric BEV Perception: A Survey [arXiv 2022; Github]
• Transformers in 3D Point Clouds: A Survey [arXiv 2022]
• Surface Reconstruction from Point Clouds: A Survey and a Benchmark [arXiv 2022]
• Sequential Point Clouds: A Survey [arXiv 2022]
• A Survey of Robust LiDAR-based 3D Object Detection Methods for Autonomous Driving [arXiv 2022]
• A Survey of Non-Rigid 3D Registration [Eurographics 2022]
• Comprehensive Review of Deep Learning-Based 3D Point Clouds Completion Processing and Analysis [TITS 2022]
• Multi-modal Sensor Fusion for Auto Driving Perception: A Survey [arXiv 2022]
• 3D Object Detection from Images for Autonomous Driving: A Survey [arXiv 2022; Github]
• Survey and Systematization of 3D Object Detection Models and Methods [arXiv 2022]
• 3D Object Detection for Autonomous Driving: A Survey [Pattern Recognition 2022; Github]
• Multi-Modal 3D Object Detection in Autonomous Driving: a Survey [arXiv 2021]
• 3D Semantic Scene Completion: a Survey [IJCV 2022]
• Deep Learning based 3D Segmentation: A Survey [arXiv 2021]
• A comprehensive survey on point cloud registration [arXiv 2021]
• Deep Learning for 3D Point Clouds: A Survey [TPAMI 2020; Github]
• A Comprehensive Performance Evaluation of 3D Local Feature Descriptors [IJCV 2016]

2023

• CVPR
  • MoDAR: Using Motion Forecasting for 3D Object Detection in Point Cloud Sequences [det]
  • TorchSparse++: Efficient Point Cloud Engine [engine; PyTorch]
  • Bi-LRFusion: Bi-Directional LiDAR-Radar Fusion for 3D Dynamic Object Detection [det; PyTorch]
  • NeurOCS: Neural NOCS Supervision for Monocular 3D Object Localization [localization]
  • GrowSP: Unsupervised Semantic Segmentation of 3D Point Clouds [seg; PyTorch]
  • 3D Registration with Maximal Cliques [registration; Github]
  • DynStatF: An Efficient Feature Fusion Strategy for LiDAR 3D Object Detection [det; CVPRW]
  • GeoMAE: Masked Geometric Target Prediction for Self-supervised Point Cloud Pre-Training [self-supervised; Github]
  • PVT-SSD: Single-Stage 3D Object Detector with Point-Voxel Transformer [det; Github]
  • SHS-Net: Learning Signed Hyper Surfaces for Oriented Normal Estimation of Point Clouds [normal estimation; PyTorch]
  • Think Twice before Driving: Towards Scalable Decoders for End-to-End Autonomous Driving [autonomous driving; PyTorch]
  • Self-supervised Pre-training with Masked Shape Prediction for 3D Scene Understanding [self-supervised]
  • PointCMP: Contrastive Mask Prediction for Self-supervised Learning on Point Cloud Videos [self-supervised]
  • PillarNeXt: Rethinking Network Designs for 3D Object Detection in LiDAR Point Clouds [det]
  • Self-Supervised 3D Scene Flow Estimation Guided by Superpoints [scene flow; Github]
  • Grad-PU: Arbitrary-Scale Point Cloud Upsampling via Gradient Descent with Learned Distance Functions [upsampling; PyTorch]
  • LiDAR2Map: In Defense of LiDAR-Based Semantic Map Construction Using Online Camera Distillation [seg; Github]
  • Density-Insensitive Unsupervised Domain Adaption on 3D Object Detection [det; PyTorch]
  • Curricular Object Manipulation in LiDAR-based Object Detection [det; PyTorch]
  • Exploiting the Complementarity of 2D and 3D Networks to Address Domain-Shift in 3D Semantic Segmentation [CVPRW]
  • Hierarchical Supervision and Shuffle Data Augmentation for 3D Semi-Supervised Object Detection [det; PyTorch]
  • IterativePFN: True Iterative Point Cloud Filtering [filtering; Github]
  • Robust Outlier Rejection for 3D Registration with Variational Bayes [registration; Github]
  • Robust Multiview Point Cloud Registration with Reliable Pose Graph Initialization and History Reweighting [registration; PyTorch]
  • 3D Semantic Segmentation in the Wild: Learning Generalized Models for Adverse-Condition Point Clouds [seg; PyTorch]
  • Understanding the Robustness of 3D Object Detection with Bird's-Eye-View Representations in Autonomous Driving [det]
  • NerVE: Neural Volumetric Edges for Parametric Curve Extraction from Point Cloud [parametric curve; Project]
  • LinK: Linear Kernel for LiDAR-based 3D Perception [det, seg; Github]
  • NeuralPCI: Spatio-temporal Neural Field for 3D Point Cloud Multi-frame Non-linear Interpolation [interpolation; Github]
  • UniDistill: A Universal Cross-Modality Knowledge Distillation Framework for 3D Object Detection in Bird's-Eye View [det; PyTorch]
  • 3D Video Object Detection with Learnable Object-Centric Global Optimization [det; Github]
  • EFEM: Equivariant Neural Field Expectation Maximization for 3D Object Segmentation Without Scene Supervision [seg; PyTorch]
  • Collaboration Helps Camera Overtake LiDAR in 3D Detection [det; Github]
  • Unsupervised Deep Probabilistic Approach for Partial Point Cloud Registration [registration; Github]
  • MonoATT: Online Monocular 3D Object Detection with Adaptive Token Transformer [det]
  • OcTr: Octree-based Transformer for 3D Object Detection [det]
  • Spherical Transformer for LiDAR-based 3D Recognition [seg, det; PyTorch]
  • Novel Class Discovery for 3D Point Cloud Semantic Segmentation [seg; PyTorch]
  • 3DQD: Generalized Deep 3D Shape Prior via Part-Discretized Diffusion Process [generation; PyTorch]
  • Self-Supervised Learning for Multimodal Non-Rigid 3D Shape Matching [non-rigid matching; PyTorch]
  • Less is More: Reducing Task and Model Complexity for 3D Point Cloud Semantic Segmentation [seg; Github]
  • CAPE: Camera View Position Embedding for Multi-View 3D Object Detection [det; Github]
  • Benchmarking Robustness of 3D Object Detection to Common Corruptions in Autonomous Driving [det; Github]
  • VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Tracking [det, tracking; PyTorch]
  • Deep Graph-based Spatial Consistency for Robust Non-rigid Point Cloud Registration [non-rigid registration; Github]
  • TBP-Former: Learning Temporal Bird's-Eye-View Pyramid for Joint Perception and Prediction in Vision-Centric Autonomous Driving [autonomous driving; Github]
  • MSeg3D: Multi-modal 3D Semantic Segmentation for Autonomous Driving [seg; PyTorch]
  • MSF: Motion-guided Sequential Fusion for Efficient 3D Object Detection from Point Cloud Sequences [det; Github]
  • Weakly Supervised Monocular 3D Object Detection using Multi-View Projection and Direction Consistency [det; Github]
  • Rotation-Invariant Transformer for Point Cloud Matching [matching]
  • Frequency-Modulated Point Cloud Rendering with Easy Editing [rendering; PyTorch]
  • Parameter is Not All You Need: Starting from Non-Parametric Networks for 3D Point Cloud Analysis [cls, seg, det; PyTorch]
  • PiMAE: Point Cloud and Image Interactive Masked Autoencoders for 3D Object Detection [det; PyTorch]
  • SCPNet: Semantic Scene Completion on Point Cloud [semantic scene completion]
  • Uni3D: A Unified Baseline for Multi-dataset 3D Object Detection [det; Github]
  • Bi3D: Bi-domain Active Learning for Cross-domain 3D Object Detection [det; Github]
  • LoGoNet: Towards Accurate 3D Object Detection with Local-to-Global Cross-Modal Fusion [det; Github]
  • X3KD: Knowledge Distillation Across Modalities, Tasks and Stages for Multi-Camera 3D Object Detection [det]
  • Virtual Sparse Convolution for Multimodal 3D Object Detection [det; PyTorch]
  • Unsupervised 3D Shape Reconstruction by Part Retrieval and Assembly [reconstruction]
  • ACL-SPC: Adaptive Closed-Loop system for Self-Supervised Point Cloud Completion [completion; Github]
  • PointCert: Point Cloud Classification with Deterministic Certified Robustness Guarantees [cls]
  • Towards Domain Generalization for Multi-view 3D Object Detection in Bird-Eye-View [det]
  • Neural Intrinsic Embedding for Non-rigid Point Cloud Matching [non-rigid matching]
  • SBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolution [seg]
  • Multimodal Industrial Anomaly Detection via Hybrid Fusion [anomaly detection; Github]
  • ProxyFormer: Proxy Alignment Assisted Point Cloud Completion with Missing Part Sensitive Transformer [completion; PyTorch]
  • Point Cloud Forecasting as a Proxy for 4D Occupancy Forecasting [autonomous driving; PyTorch]
  • VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene Completion [autonomous driving; Github]
  • Tri-Perspective View for Vision-Based 3D Semantic Occupancy Prediction [autonomous driving; PyTorch]
  • CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP [pre-training]
  • ConQueR: Query Contrast Voxel-DETR for 3D Object Detection [det; Github]
  • Learning 3D Representations from 2D Pre-trained Models via Image-to-Point Masked Autoencoders [pre-training; Github]
  • GD-MAE: Generative Decoder for MAE Pre-training on LiDAR Point Clouds [pre-training; Github]
  • BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View Recognition via Perspective Supervision [autonomous driving]
  • MSMDFusion: Fusing LiDAR and Camera at Multiple Scales with Multi-Depth Seeds for 3D Object Detection [det; Github]
  • DETRs with Hybrid Matching [det; Github]
  • LaserMix for Semi-Supervised LiDAR Semantic Segmentation [seg; PyTorch]
  • PointDistiller: Structured Knowledge Distillation Towards Efficient and Compact 3D Detection [det; Github]
• AAAI
  • Context-Aware Transformer for 3D Point Cloud Automatic Annotation [annotation]
  • GAM : Gradient Attention Module of Optimization for Point Clouds Analysis [seg, det, cls; PyTorch]
  • Parametric Surface Constrained Upsampler Network for Point Cloud [upsampling; PyTorch]
  • CRIN: Rotation-Invariant Point Cloud Analysis and Rotation Estimation via Centrifugal Reference Frame [rotation invariance]
  • PUPS: Point Cloud Unified Panoptic Segmentation [seg]
  • StereoDistill: Pick the Cream from LiDAR for Distilling Stereo-based 3D Object Detection [det]
  • Rethinking Rotation Invariance with Point Cloud Registration [cls, seg, retrieval; PyTorch]
  • SSDA3D: Semi-supervised Domain Adaptation for 3D Object Detection from Point Cloud [det; Github]
  • CL3D: Unsupervised Domain Adaptation for Cross-LiDAR 3D Detection [det; Github]
  • MGTANet: Encoding Sequential LiDAR Points Using Long Short-Term Motion-Guided Temporal Attention for 3D Object Detection [det; PyTorch]
  • Attention-based Depth Distillation with 3D-Aware Positional Encoding for Monocular 3D Object Detection [det; Github]
  • NeAF: Learning Neural Angle Fields for Point Normal Estimation [normal estimation; PyTorch]
  • Superpoint Transformer for 3D Scene Instance Segmentation [seg; PyTorch]
  • Language-Assisted 3D Feature Learning for Semantic Scene Understanding [det, seg; Github]
  • CasFusionNet: A Cascaded Network for Point Cloud Semantic Scene Completion by Dense Feature Fusion [semantic scene completion; Github]
  • Transformation-Equivariant 3D Object Detection for Autonomous Driving [det]
  • PointCA: Evaluating the Robustness of 3D Point Cloud Completion Models Against Adversarial Examples [completion]
  • GLT-T: Global-Local Transformer Voting for 3D Single Object Tracking in Point Clouds [tracking; Github]
  • Normal Transformer: Extracting Surface Geometry from LiDAR Points Enhanced by Visual Semantics [normal estimation]
  • A Simple Baseline for Multi-Camera 3D Object Detection [det; Github]
  • PolarFormer: Multi-camera 3D Object Detection with Polar Transformer [det; PyTorch]
  • BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detection [det; PyTorch]
  • BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection with Dynamic Temporal Stereo [det; PyTorch]
• Others
  • Energy-based Detection of Adverse Weather Effects in LiDAR Data [autonomous driving; Github; RAL]
  • Prototype Adaption and Projection for Few- and Zero-shot 3D Point Cloud Semantic Segmentation [seg; Github; TIP]
  • Cross-source Point Cloud Registration: Challenges, Progress and Prospects [registration; Neurocomputing]
  • OctFormer: Octree-based Transformers for 3D Point Clouds [seg, det; Github; TOG]
  • APR: Online Distant Point Cloud Registration Through Aggregated Point Cloud Reconstruction [registration; Github; IJCAI]
  • OSP2B: One-Stage Point-to-Box Network for 3D Siamese Tracking [tracking; IJCAI]
  • A Closer Look at Few-Shot 3D Point Cloud Classification [cls; Github; IJCV]
  • HybridPoint: Point Cloud Registration Based on Hybrid Point Sampling and Matching [registration; Github; ICME]
  • Variational Relational Point Completion Network for Robust 3D Classification [completion; TPAMI]
  • RoReg: Pairwise Point Cloud Registration with Oriented Descriptors and Local Rotations [registration; PyTorch; TPAMI]
  • Analogy-Forming Transformers for Few-Shot 3D Parsing [seg; Project; ICLR]
  • Joint 2D-3D multi-task scene understanding on Cityscapes-3D: 3D detection, segmentation, and depth estimation [seg, det, depth estimation; PyTorch; ICLR]
  • CircNet: Meshing 3D Point Clouds with Circumcenter Detection [triangulation; Github; ICLR]
  • BEVDistill: Cross-Modal BEV Distillation for Multi-View 3D Object Detection [det; Github; ICLR]
  • Bidirectional Propagation for Cross-Modal 3D Object Detection [det; PyTorch; ICLR]
  • Exploring Active 3D Object Detection from a Generalization Perspective [det; PyTorch; ICLR]
  • DDS3D: Dense Pseudo-Labels with Dynamic Threshold for Semi-Supervised 3D Object Detection [det; Github; ICRA]
  • Are All Point Clouds Suitable for Completion? Weakly Supervised Quality Evaluation Network for Point Cloud Completion [completion; ICRA]
  • A Unified BEV Model for Joint Learning of 3D Local Features and Overlap Estimation [registration; ICRA]
  • LODE: Locally Conditioned Eikonal Implicit Scene Completion from Sparse LiDAR [completion; PyTorch; ICRA]
  • DuEqNet: Dual-Equivariance Network in Outdoor 3D Object Detection for Autonomous Driving [det; ICRA]
  • Few-Shot Point Cloud Semantic Segmentation via Contrastive Self-Supervision and Multi-Resolution Attention [seg; ICRA]
  • MVFusion: Multi-View 3D Object Detection with Semantic-aligned Radar and Camera Fusion [det; ICRA]
  • MonoPGC: Monocular 3D Object Detection with Pixel Geometry Contexts [det; ICRA]
  • SCARP: 3D Shape Completion in ARbitrary Poses for Improved Grasping [completion; PyTorch; ICRA]
  • BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation [autonomous driving; PyTorch; ICRA]
  • CrossDTR: Cross-view and Depth-guided Transformers for 3D Object Detection [det; Github; ICRA]
  • ConDA: Unsupervised Domain Adaptation for LiDAR Segmentation via Regularized Domain Concatenation [seg, unsupervised domain adaptation; Github; ICRA]
  • MonoEdge: Monocular 3D Object Detection Using Local Perspectives [det; WACV]
  • Far3Det: Towards Far-Field 3D Detection [det; WACV]
  • Centroid Distance Keypoint Detector for Colored Point Clouds [keypoint; Github; WACV]
  • NeuralBF: Neural Bilateral Filtering for Top-down Instance Segmentation on Point Clouds [seg; Project; WACV]
• arXiv
  • SAM3D: Segment Anything in 3D Scenes [seg; PyTorch]
  • SAM3D: Zero-Shot 3D Object Detection via Segment Anything Model [det; PyTorch]
  • Scene as Occupancy [autonomous driving; PyTorch]
  • Collect-and-Distribute Transformer for 3D Point Cloud Analysis [cls, seg; PyTorch]
  • OCBEV: Object-Centric BEV Transformer for Multi-View 3D Object Detection [det]
  • AD-PT: Autonomous Driving Pre-Training with Large-scale Point Cloud Dataset [pre-training; PyTorch]
  • Point-GCC: Universal Self-supervised 3D Scene Pre-training via Geometry-Color Contrast [self-supervised learning; Github]
  • Occ-BEV: Multi-Camera Unified Pre-training via 3D Scene Reconstruction [det, semantic occupancy prediction; PyTorch]
  • BEV-IO: Enhancing Bird's-Eye-View 3D Detection with Instance Occupancy [det]
  • Banana: Banach Fixed-Point Network for Pointcloud Segmentation with Inter-Part Equivariance [seg]
  • All Points Matter: Entropy-Regularized Distribution Alignment for Weakly-supervised 3D Segmentation [seg]
  • Learning Occupancy for Monocular 3D Object Detection [det; Github]
  • SAD: Segment Any RGBD [seg; PyTorch]
  • Real-Aug: Realistic Scene Synthesis for LiDAR Augmentation in 3D Object Detection [augmentation]
  • PointGPT: Auto-regressively Generative Pre-training from Point Clouds [pre-training; Github]
  • SUG: Single-dataset Unified Generalization for 3D Point Cloud Classification [cls; PyTorch]
  • Multi-Modal 3D Object Detection by Box Matching [det; Github]
  • DSPDet3D: Dynamic Spatial Pruning for 3D Small Object Detection [det; PyTorch]
  • SparseFusion: Fusing Multi-Modal Sparse Representations for Multi-Sensor 3D Object Detection [det; Github]
  • DQS3D: Densely-matched Quantization-aware Semi-supervised 3D Detection [det; PyTorch]
  • ContrastMotion: Self-supervised Scene Motion Learning for Large-Scale LiDAR Point Clouds [scene motion]
  • Segment Anything in 3D with NeRFs [seg; Github]
  • You Never Get a Second Chance To Make a Good First Impression: Seeding Active Learning for 3D Semantic Segmentation [annotation; PyTorch]
  • Fully Sparse Fusion for 3D Object Detection [det; Github]
  • Once Detected, Never Lost: Surpassing Human Performance in Offline LiDAR based 3D Object Detection [det; PyTorch]
  • MetaBEV: Solving Sensor Failures for BEV Detection and Map Segmentation [det, seg; Github]
  • 3D Feature Prediction for Masked-AutoEncoder-Based Point Cloud Pretraining [pre-training]
  • Swin3D: A Pretrained Transformer Backbone for 3D Indoor Scene Understanding [seg, det; Github]
  • RoboBEV: Towards Robust Bird's Eye View Perception under Corruptions [autonomous driving; Github]
  • OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy Prediction [semantic occupancy prediction; PyTorch]
  • Sparse Dense Fusion for 3D Object Detection [det]
  • BEVStereo++: Accurate Depth Estimation in Multi-view 3D Object Detection via Dynamic Temporal Stereo [det]
  • VPFusion: Towards Robust Vertical Representation Learning for 3D Object Detection [det]
  • Geometric-aware Pretraining for Vision-centric 3D Object Detection [det; PyTorch]
  • MS3D: Leveraging Multiple Detectors for Unsupervised Domain Adaptation in 3D Object Detection [det; PyTorch]
  • Temporal Enhanced Training of Multi-view 3D Object Detector via Historical Object Prediction [det; Github]
  • VoxelFormer: Bird's-Eye-View Feature Generation based on Dual-view Attention for Multi-view 3D Object Detection [det; PyTorch]
  • EA-BEV: Edge-aware Bird' s-Eye-View Projector for 3D Object Detection [det; Github]
  • IC-FPS: Instance-Centroid Faster Point Sampling Module for 3D Point-base Object Detection [det; Github]
  • APPT : Asymmetric Parallel Point Transformer for 3D Point Cloud Understanding [seg, cls]
  • BEVFusion4D: Learning LiDAR-Camera Fusion Under Bird's-Eye-View via Cross-Modality Guidance and Temporal Aggregation [det]
  • Point2Pix: Photo-Realistic Point Cloud Rendering via Neural Radiance Fields [rendering]
  • Point2Vec for Self-Supervised Representation Learning on Point Clouds [self-supervised; Project]
  • DORT: Modeling Dynamic Objects in Recurrent for Multi-Camera 3D Object Detection and Tracking [det, tracking; Github]
  • BEVSimDet: Simulated Multi-modal Distillation in Bird's-Eye View for Multi-view 3D Object Detection [det; Github]
  • UnScene3D: Unsupervised 3D Instance Segmentation for Indoor Scenes [seg; Project]
  • Robo3D: Towards Robust and Reliable 3D Perception against Corruptions [seg, det; Github]
  • ByteTrackV2: 2D and 3D Multi-Object Tracking by Associating Every Detection Box [tracking; Github]
  • StereoScene: BEV-Assisted Stereo Matching Empowers 3D Semantic Scene Completion [semantic scene completion; PyTorch]
  • Position-Guided Point Cloud Panoptic Segmentation Transformer [seg; Github]
  • CLIP^2: Contrastive Language-Image-Point Pretraining from Real-World Point Cloud Data [pre-training]
  • An Effective Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds [tracking; PyTorch]
  • LiDARFormer: A Unified Transformer-based Multi-task Network for LiDAR Perception [det, seg]
  • CurveCloudNet: Processing Point Clouds with 1D Structure [seg]
  • Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D Object Detection [det; Github]
  • 3D Data Augmentation for Driving Scenes on Camera [autonomous driving]
  • Vehicle-Infrastructure Cooperative 3D Object Detection via Feature Flow Prediction [autonomous driving; PyTorch]
  • A Simple Attempt for 3D Occupancy Estimation in Autonomous Driving [autonomous driving; Github]
  • SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Driving [autonomous driving; PyTorch]
  • DiffBEV: Conditional Diffusion Model for Bird's Eye View Perception [seg, det; PyTorch]
  • GeoSpark: Sparking up Point Cloud Segmentation with Geometry Clue [seg]
  • BEVHeight: A Robust Framework for Vision-based Roadside 3D Object Detection [det; PyTorch]
  • Modeling Continuous Motion for 3D Point Cloud Object Tracking [tracking]
  • Category-Level Multi-Part Multi-Joint 3D Shape Assembly [shape assembly]
  • Exploring Recurrent Long-term Temporal Fusion for Multi-view 3D Perception [autonomous driving]
  • Rethinking Range View Representation for LiDAR Segmentation [seg]
  • Efficient Transformer-based 3D Object Detection with Dynamic Token Halting [det]
  • ElC-OIS: Ellipsoidal Clustering for Open-World Instance Segmentation on LiDAR Data [seg; Github]
  • InsMOS: Instance-Aware Moving Object Segmentation in LiDAR Data [seg; Github]
  • Joint-MAE: 2D-3D Joint Masked Autoencoders for 3D Point Cloud Pre-training [pre-training]
  • OccDepth: A Depth-Aware Method for 3D Semantic Scene Completion [autonomous driving; PyTorch]
  • DA-BEV: Depth Aware BEV Transformer for 3D Object Detection [det]
  • Pillar R-CNN for Point Cloud 3D Object Detection [det]
  • 2D-Empowered 3D Object Detection on the Edge [det]
  • General Rotation Invariance Learning for Point Clouds via Weight-Feature Alignment [cls, seg, det]
  • PointWavelet: Learning in Spectral Domain for 3D Point Cloud Analysis [cls, seg]
  • Standing Between Past and Future: Spatio-Temporal Modeling for Multi-Camera 3D Multi-Object Tracking [tracking; Github]
  • Top-Down Beats Bottom-Up in 3D Instance Segmentation [seg; PyTorch]
  • TR3D: Towards Real-Time Indoor 3D Object Detection [det; PyTorch]
  • CVTNet: A Cross-View Transformer Network for Place Recognition Using LiDAR Data [place recognition; Github]
  • AOP-Net: All-in-One Perception Network for Joint LiDAR-based 3D Object Detection and Panoptic Segmentation [det, seg]
  • Fast-BEV: A Fast and Strong Bird’s-Eye View Perception Baseline [det; PyTorch]
  • On the Adversarial Robustness of Camera-based 3D Object Detection [det]
  • RangeViT: Towards Vision Transformers for 3D Semantic Segmentation in Autonomous Driving [seg; Github]
  • FlatFormer: Flattened Window Attention for Efficient Point Cloud Transformer [det]
  • Joint Representation Learning for Text and 3D Point Cloud [pre-training; Github]
  • Long Range Pooling for 3D Large-Scale Scene Understanding [seg]
  • OA-BEV: Bringing Object Awareness to Bird’s-Eye-View Representation for Multi-Camera 3D Object Detection [det]
  • SAT: Size-Aware Transformer for 3D Point Cloud Semantic Segmentation [seg]
  • FrustumFormer: Adaptive Instance-aware Resampling for Multi-view 3D Detection [det]
  • AdaPoinTr: Diverse Point Cloud Completion with Adaptive Geometry-Aware Transformers [completion]
  • Anchor3DLane: Learning to Regress 3D Anchors for Monocular 3D Lane Detection [lane det]
  • Object as Query: Equipping Any 2D Object Detector with 3D Detection Ability [det]
  • Super Sparse 3D Object Detection [det; PyTorch]
  • Model-Agnostic Hierarchical Attention for 3D Object Detection [det; Github]
  • Ponder: Point Cloud Pre-training via Neural Rendering [self-supervised learning; Github]
  • Cross Modal Transformer via Coordinates Encoding for 3D Object Dectection [det; Github]
  • Benchmarking the Robustness of LiDAR Semantic Segmentation Models [seg]

2022

• ECCV
  • PCR-CG: Point Cloud Registration via Color and Geometry [registration; PyTorch]
  • SPE-Net: Boosting Point Cloud Analysis via Rotation Robustness Enhancement [cls, seg; PyTorch]
  • Cross-Modality Knowledge Distillation Network for Monocular 3D Object Detection [det; PyTorch]
  • LiDAL: Inter-frame Uncertainty Based Active Learning for 3D LiDAR Semantic Segmentation [seg; Github]
  • Homogeneous Multi-modal Feature Fusion and Interaction for 3D Object Detection [det]
  • Improving the Intra-class Long-tail in 3D Detection via Rare Example Mining [det]
  • CramNet: Camera-Radar Fusion with Ray-Constrained Cross-Attention for Robust 3D Object Detection [det]
  • LESS: Label-Efficient Semantic Segmentation for LiDAR Point Clouds [seg]
  • Autoregressive Uncertainty Modeling for 3D Bounding Box Prediction [det; PyTorch]
  • SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds [det]
  • LidarNAS: Unifying and Searching Neural Architectures for 3D Point Clouds [NAS]
  • FBNet: Feedback Network for Point Cloud Completion [completion; PyTorch]
  • INT: Towards Infinite-frames 3D Detection with An Efficient Framework [det]
  • MPPNet: Multi-Frame Feature Intertwining with Proxy Points for 3D Temporal Object Detection [det; Github]
  • PointInst3D: Segmenting 3D Instances by Points [seg; Github]
  • PillarNet: Real-Time and High-Performance Pillar-based 3D Object Detection [det; PyTorch]
  • GitNet: Geometric Prior-based Transformation for Birds-Eye-View Segmentation [autonomous driving]
  • Geodesic-Former: a Geodesic-Guided Few-shot 3D Point Cloud Instance Segmenter [seg; Github]
  • Multimodal Transformer for Automatic 3D Annotation and Object Detection [det; PyTorch]
  • Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using Bounding Boxes [seg; PyTorch]
  • Online Segmentation of LiDAR Sequences: Dataset and Algorithm [seg; PyTorch]
  • SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse views [reconstruction; PyTorch]
  • Language-Grounded Indoor 3D Semantic Segmentation in the Wild [seg; PyTorch]
  • CenterFormer: Center-based Transformer for 3D Object Detection [det; Github]
  • Learning to Generate Realistic LiDAR Point Clouds [generation; Project]
  • PointCLM: A Contrastive Learning-based Framework for Multi-instance Point Cloud Registration [registration; Github]
  • SimpleRecon: 3D Reconstruction Without 3D Convolutions [reconstruction; Project]
  • Improving RGB-D Point Cloud Registration by Learning Multi-scale Local Linear Transformation [registration; Github]
  • Masked Discrimination for Self-Supervised Learning on Point Clouds [self-supervised; PyTorch]
  • MORE: Multi-Order RElation Mining for Dense Captioning in 3D Scenes [dense captioning; Github]
  • Objects Can Move: 3D Change Detection by Geometric Transformation Constistency [change detection; Github]
  • PointTree: Transformation-Robust Point Cloud Encoder with Relaxed K-D Trees [cls, seg; PyTorch]
  • diffConv: Analyzing Irregular Point Clouds with an Irregular View [cls, seg; PyTorch]
  • SLiDE: Self-supervised LiDAR De-snowing through Reconstruction Difficulty [de-snowing]
  • Graph R-CNN: Towards Accurate 3D Object Detection with Semantic-Decorated Local Graph [det; PyTorch]
  • Point Primitive Transformer for Long-Term 4D Point Cloud Video Understanding [4D]
  • Few-shot Single-view 3D Reconstruction with Memory Prior Contrastive Network [reconstruction]
  • PolarMOT: How Far Can Geometric Relations Take Us in 3D Multi-Object Tracking? [tracking]
  • MonteBoxFinder: Detecting and Filtering Primitives to Fit a Noisy Point Cloud [representation]
  • Large-displacement 3D Object Tracking with Hybrid Non-local Optimization [tracking; Github]
  • ProposalContrast: Unsupervised Pre-training for LiDAR-based 3D Object Detection [det; Github]
  • Semi-supervised 3D Object Detection with Proficient Teachers [det; Github]
  • Monocular 3D Object Detection with Depth from Motion [det; PyTorch]
  • GraphFit: Learning Multi-scale Graph-Convolutional Representation for Point Cloud Normal Estimation [normal estimation; PyTorch]
  • PatchRD: Detail-Preserving Shape Completion by Learning Patch Retrieval and Deformation [completion; PyTorch]
  • Label-Guided Auxiliary Training Improves 3D Object Detector [det; Github]
  • Salient Object Detection for Point Clouds [det; Code]
  • 3D Siamese Transformer Network for Single Object Tracking on Point Clouds [tracking; Github]
  • Point Cloud Compression with Sibling Context and Surface Priors [compression; PyTorch]
  • SQN: Weakly-Supervised Semantic Segmentation of Large-Scale 3D Point Clouds with 1000× Fewer Labels [seg; Tensorflow]
  • PointMixer: MLP-Mixer for Point Cloud Understanding [seg, cls, reconstruction; PyTorch]
  • DEVIANT: Depth EquiVarIAnt NeTwork for Monocular 3D Object Detection [det, monocular; PyTorch]
  • Dynamic 3D Scene Analysis by Point Cloud Accumulation [accumulation; PyTorch]
  • MeshMAE: Masked Autoencoders for 3D Mesh Data Analysis [self-supervised]
  • SeedFormer: Patch Seeds based Point Cloud Completion with Upsample Transformer [completion; PyTorch]
  • Unsupervised Deep Multi-Shape Matching [matching]
  • Monocular 3D Object Reconstruction with GAN Inversion [reconstruction; PyTorch]
  • CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentation [seg; PyTorch]
  • GIPSO: Geometrically Informed Propagation for Online Adaptation in 3D LiDAR Segmentation [seg; PyTorch]
  • Densely Constrained Depth Estimator for Monocular 3D Object Detection [det, monocular; PyTorch]
  • SuperLine3D: Self-supervised Line Segmentation and Description for LiDAR Point Cloud [registration; Tensorflow]
  • What Matters for 3D Scene Flow Network [scene flow; PyTorch]
  • Dual Adaptive Transformations for Weakly Supervised Point Cloud Segmentation [seg]
  • Towards High-Fidelity Single-view Holistic Reconstruction of Indoor Scenes [reconstruction; Github]
  • JPerceiver: Joint Perception Network for Depth, Pose and Layout Estimation in Driving Scenes [autonomous driving; PyTorch]
  • TransGrasp: Grasp Pose Estimation of a Category of Objects by Transferring Grasps from Only One Labeled Instance [pose estimation; PyTorch]
  • CATRE: Iterative Point Clouds Alignment for Category-level Object Pose Refinement [pose estimation; PyTorch]
  • Lidar Point Cloud Guided Monocular 3D Object Detection [det, monocular; PyTorch]
  • DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection [det, monocular; PyTorch]
  • ST-P3: End-to-end Vision-based Autonomous Driving via Spatial-Temporal Feature Learning [autonomous driving; PyTorch]
  • 3D Instances as 1D Kernels [seg; PyTorch]
  • Bi-PointFlowNet: Bidirectional Learning for Point Cloud Based Scene Flow Estimation [scene flow; PyTorch]
  • Rethinking IoU-based Optimization for Single-stage 3D Object Detection [det; Github]
  • Meta-Sampler: Almost-Universal yet Task-Oriented Sampling for Point Clouds [sampling]
  • CPO: Change Robust Panorama to Point Cloud Localization [visual localization]
  • Category-Level 6D Object Pose and Size Estimation using Self-Supervised Deep Prior Deformation Networks [pose estimation; PyTorch]
  • Hierarchical Latent Structure for Multi-Modal Vehicle Trajectory Forecasting [autonomous driving; PyTorch]
  • 2DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Clouds [seg; Github]
  • A Closer Look at Invariances in Self-supervised Pre-training for 3D Vision [self-supervised]
  • Open-world Semantic Segmentation for LIDAR Point Clouds [seg; PyTorch]
  • BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers [det, seg; Github]
  • LiDAR Distillation: Bridging the Beam-Induced Domain Gap for 3D Object Detection [det; PyTorch]
  • FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection [det; PyTorch]
  • Masked Autoencoders for Point Cloud Self-supervised Learning [self-supervised; PyTorch]
  • PETR: Position Embedding Transformation for Multi-View 3D Object Detection [det; PyTorch]
  • Learning Ego 3D Representation as Ray Tracing [autonomous driving; PyTorch]
  • AutoAlignV2: Deformable Feature Aggregation for Dynamic Multi-Modal 3D Object Detection [det; Github]
• CVPR
  • Bridged Transformer for Vision and Point Cloud 3D Object Detection [det]
  • ShapeFormer: Transformer-based Shape Completion via Sparse Representation [completion; PyTorch]
  • RBGNet: Ray-based Grouping for 3D Object Detection [det; Github]
  • Boosting 3D Object Detection by Simulating Multimodality on Point Clouds [det]
  • MonoGround: Detecting Monocular 3D Objects from the Ground [det, monocular; Github]
  • PlanarRecon: Real-time 3D Plane Detection and Reconstruction from Posed Monocular Videos [reconstruction; PyTorch]
  • Learning 3D Object Shape and Layout without 3D Supervision [shape, layout; Project]
  • Deterministic Point Cloud Registration via Novel Transformation Decomposition [registration]
  • Cross-view Transformers for real-time Map-view Semantic Segmentation [autonomous driving; PyTorch]
  • RIDDLE: Lidar Data Compression with Range Image Deep Delta Encoding [compression]
  • Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentation [seg; Github]
  • Voxel Field Fusion for 3D Object Detection [det; PyTorch]
  • On the Choice of Data for Efficient Training and Validation of End-to-End Driving Models [autonomous driving; CVPRW]
  • 3PSDF: Three-Pole Signed Distance Function for Learning Surfaces with Arbitrary Topologies [reconstruction]
  • The Devil is in the Pose: Ambiguity-free 3D Rotation-invariant Learning via Pose-aware Convolution [cls, seg]
  • Neural Shape Mating: Self-Supervised Object Assembly with Adversarial Shape Priors [shape mating; Project]
  • Time3D: End-to-End Joint Monocular 3D Object Detection and Tracking for Autonomous Driving [autonomous driving, monocular]
  • SemAffiNet: Semantic-Affine Transformation for Point Cloud Segmentation [seg; PyTorch]
  • Primitive3D: 3D Object Dataset Synthesis from Randomly Assembled Primitives [cls]
  • RCP: Recurrent Closest Point for Scene Flow Estimation on 3D Point Clouds [scene flow; Github]
  • Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection [det]
  • Surface Representation for Point Clouds [cls, seg, det; PyTorch]
  • FvOR: Robust Joint Shape and Pose Optimization for Few-view Object Reconstruction [reconstruction; PyTorch]
  • Topologically-Aware Deformation Fields for Single-View 3D Reconstruction [reconstruction; Project]
  • Deep Orientation-Aware Functional Maps: Tackling Symmetry Issues in Shape Matching [matching; PyTorch]
  • Rotationally Equivariant 3D Object Detection [det; Project]
  • MM-TTA: Multi-Modal Test-Time Adaptation for 3D Semantic Segmentation [seg; Project]
  • Density-preserving Deep Point Cloud Compression [compression; PyTorch]
  • Coupled Iterative Refinement for 6D Multi-Object Pose Estimation [pose estimation; PyTorch]
  • A Scalable Combinatorial Solver for Elastic Geometrically Consistent 3D Shape Matching [match; Github]
  • Focal Sparse Convolutional Networks for 3D Object Detection [det; PyTorch]
  • Surface Reconstruction from Point Clouds by Learning Predictive Context Priors [reconstruction; Tensorflow]
  • Reconstructing Surfaces for Sparse Point Clouds with On-Surface Priors [reconstruction; Tensorflow]
  • Forecasting from LiDAR via Future Object Detection [forecasting; PyTorch]
  • Fast Point Transformer [seg, det; PyTorch]
  • Proposal-free Lidar Panoptic Segmentation with Pillar-level Affinity [seg; CVPRW]
  • Multi-Camera Multiple 3D Object Tracking on the Move for Autonomous Vehicles [autonomous driving; CVPRW]
  • Learning Multi-View Aggregation In the Wild for Large-Scale 3D Semantic Segmentation [seg; PyTorch]
  • OccAM's Laser: Occlusion-based Attribution Maps for 3D Object Detectors on LiDAR Data [det; PyTorch]
  • 3DeformRS: Certifying Spatial Deformations on Point Clouds [robustness; Github]
  • HyperDet3D: Learning a Scene-conditioned 3D Object Detector [det]
  • Exploiting Temporal Relations on Radar Perception for Autonomous Driving [autonomous driving]
  • Homography Loss for Monocular 3D Object Detection [det]
  • CAT-Det: Contrastively Augmented Transformer for Multi-modal 3D Object Detection [det]
  • Learning to Detect Mobile Objects from LiDAR Scans Without Labels [det; PyTorch]
  • Learning Local Displacements for Point Cloud Completion [completion]
  • Deformation and Correspondence Aware Unsupervised Synthetic-to-Real Scene Flow Estimation for Point Clouds [scene flow; Github]
  • LiDAR Snowfall Simulation for Robust 3D Object Detection [det; Github]
  • Text2Pos: Text-to-Point-Cloud Cross-Modal Localization [localization; PyTorch]
  • Stratified Transformer for 3D Point Cloud Segmentation [seg; PyTorch]
  • REGTR: End-to-end Point Cloud Correspondences with Transformers [registration; PyTorch]
  • Equivariant Point Cloud Analysis via Learning Orientations for Message Passing [cls, seg, normal estimation; Github]
  • SC^2-PCR: A Second Order Spatial Compatibility for Efficient and Robust Point Cloud Registration [registration; Github]
  • Towards Implicit Text-Guided 3D Shape Generation [generation; PyTorch]
  • Point2Seq: Detecting 3D Objects as Sequences [det; PyTorch]
  • MonoDETR: Depth-aware Transformer for Monocular 3D Object Detection [det, monocular; Github]
  • AziNorm: Exploiting the Radial Symmetry of Point Cloud for Azimuth-Normalized 3D Perception [det, seg; Github]
  • IDEA-Net: Dynamic 3D Point Cloud Interpolation via Deep Embedding Alignment [interpolation; Github]
  • TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with Transformers [det; PyTorch]
  • Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds [det; PyTorch]
  • No Pain, Big Gain: Classify Dynamic Point Cloud Sequences with Static Models by Fitting Feature-level Space-time Surfaces [cls; Github]
  • MonoDTR: Monocular 3D Object Detection with Depth-Aware Transformer [det, monocular; Github]
  • Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Clouds [det; PyTorch]
  • VISTA: Boosting 3D Object Detection via Dual Cross-VIew SpaTial Attention [det; PyTorch]
  • Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completion [det]
  • AutoSDF: Shape Priors for 3D Completion, Reconstruction and Generation [completion, reconstruction, generation]
  • DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection [det; Tensorflow]
  • Scribble-Supervised LiDAR Semantic Segmentation [seg; Github]
  • MonoJSG: Joint Semantic and Geometric Cost Volume for Monocular 3D Object Detection [det, monocular; Github]
  • PTTR: Relational 3D Point Cloud Object Tracking with Transformer [tracking; PyTorch]
  • AutoGPart: Intermediate Supervision Search for Generalizable 3D Part Segmentation [seg]
  • Point Density-Aware Voxels for LiDAR 3D Object Detection [det; Github]
  • Contrastive Boundary Learning for Point Cloud Segmentation [seg; Github]
  • Back to Reality: Weakly-supervised 3D Object Detection with Shape-guided Label Enhancement [det; PyTorch]
  • Shape-invariant 3D Adversarial Point Clouds [adversarial; Github]
  • Iterative Corresponding Geometry: Fusing Region and Depth for Highly Efficient 3D Tracking of Textureless Objects [tracking; Github]
  • ART-Point: Improving Rotation Robustness of Point Cloud Classifiers via Adversarial Rotation [cls; Github]
  • Geometric Transformer for Fast and Robust Point Cloud Registration [registration; PyTorch]
  • Lepard: Learning partial point cloud matching in rigid and deformable scenes [registration; PyTorch]
  • Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving [det, monocular; Github]
  • A Versatile Multi-View Framework for LiDAR-based 3D Object Detection with Guidance from Panoptic Segmentation [det]
  • Embracing Single Stride 3D Object Detector with Sparse Transformer [det; PyTorch]
  • Canonical Voting: Towards Robust Oriented Bounding Box Detection in 3D Scenes [det; PyTorch]
  • CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understanding [cross-modal learning; PyTorch]
  • PointCLIP: Point Cloud Understanding by CLIP [cross-modal learning; Github]
  • SoftGroup for 3D Instance Segmentation on Point Clouds [seg; PyTorch]
  • Beyond 3D Siamese Tracking: A Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds [tracking; PyTorch]
  • A Unified Query-based Paradigm for Point Cloud Understanding [det, seg, cls]
• AAAI
  • SRCN3D: Sparse R-CNN 3D Surround-View Camera Object Detection and Tracking for Autonomous Driving [det, tracking; PyTorch]
  • Static-Dynamic Co-Teaching for Class-Incremental 3D Object Detection [det; PyTorch]
  • Semantic Segmentation for Point Cloud Scenes via Dilated Graph Feature Aggregation and Pyramid Decoders [seg]
  • SASA: Semantics-Augmented Set Abstraction for Point-based 3D Object Detection [det; PyTorch]
  • DuMLP-Pin: A Dual-MLP-dot-product Permutation-invariant Network for Set Feature Extraction [cls, seg; PyTorch]
  • Reliable Inlier Evaluation for Unsupervised Point Cloud Registration [registration; Github]
  • Sparse Cross-scale Attention Network for Efficient LiDAR Panoptic Segmentation [seg]
  • FINet: Dual Branches Feature Interaction for Partial-to-Partial Point Cloud Registration [registration; Github]
  • DetarNet: Decoupling Translation and Rotation by Siamese Network for Point Cloud Registration [registration; Github]
  • AFDetV2: Rethinking the Necessity of the Second Stage for Object Detection from Point Clouds [det]
  • Joint 3D Object Detection and Tracking Using Spatio-Temporal Representation of Camera Image and LiDAR Point Clouds [det, tracking]
  • Attention-based Transformation from Latent Features to Point Clouds [generation]
  • Behind the Curtain: Learning Occluded Shapes for 3D Object Detection [det; PyTorch]
• Others
  • MoGDE: Boosting Mobile Monocular 3D Object Detection with Ground Depth Estimation [det; NeurIPS]
  • Fast-BEV: Towards Real-time On-vehicle Bird’s-Eye View Perception [det; Github; NeurIPSW]
  • Sparse2Dense: Learning to Densify 3D Features for 3D Object Detection [det; Github; NeurIPS]
  • Language Conditioned Spatial Relation Reasoning for 3D Object Grounding [localizing; PyTorch; NeurIPS]
  • A Benchmark for Out of Distribution Detection in Point Cloud 3D Semantic Segmentation [seg; NeurIPSW]
  • Analyzing Deep Learning Representations of Point Clouds for Real-Time In-Vehicle LiDAR Perception [autonomous driving; NeurIPSW]
  • HSurf-Net: Normal Estimation for 3D Point Clouds by Learning Hyper Surfaces [normal estimation; Github; NeurIPS]
  • LION: Latent Point Diffusion Models for 3D Shape Generation [generation; Github; NeurIPS]
  • SageMix: Saliency-Guided Mixup for Point Clouds [augmentation; Github; NeurIPS]
  • Prototypical VoteNet for Few-Shot 3D Point Cloud Object Detection [det; Github; NeurIPS]
  • Point Transformer V2: Grouped Vector Attention and Partition-based Pooling [cls, seg; Github; NeurIPS]
  • Let Images Give You More:Point Cloud Cross-Modal Training for Shape Analysis [cross-modality; Github; NeurIPS]
  • OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Clouds [seg; PyTorch; NeurIPS]
  • CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds [det; Github; NeurIPS]
  • Learning Consistency-Aware Unsigned Distance Functions Progressively from Raw Point Clouds [reconstruction; PyTorch; NeurIPS]
  • PolarMix: A General Data Augmentation Technique for LiDAR Point Clouds [augmentation; PyTorch; NeurIPS]
  • Fully Convolutional One-Stage 3D Object Detection on LiDAR Range Images [det; NeurIPS]
  • P2P: Tuning Pre-trained Image Models for Point Cloud Analysis with Point-to-Pixel Prompting [pre-training; PyTorch; NeurIPS]
  • Fully Sparse 3D Object Detection [det; PyTorch; NeurIPS]
  • PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies [seg, cls; PyTorch; NeurIPS]
  • K-Radar: 4D Radar Object Detection Dataset and Benchmark for Autonomous Driving in Various Weather Conditions [autonomous driving; Github; NeurIPS Track]
  • SNAKE: Shape-aware Neural 3D Keypoint Field [keypoints; PyTorch; NeurIPS]
  • Unifying Voxel-based Representation with Transformer for 3D Object Detection [det; PyTorch; NeurIPS]
  • Non-rigid Point Cloud Registration with Neural Deformation Pyramid [non-rigid, registration; PyTorch; NeurIPS]
  • DeepInteraction: 3D Object Detection via Modality Interaction [det; Github; NeurIPS]
  • Spatial Pruned Sparse Convolution for Efficient 3D Object Detection [det; NeurIPS]
  • Rethinking the compositionality of point clouds through regularization in the hyperbolic space [cls; Github; NeurIPS]
  • Cross-modal Learning for Image-Guided Point Cloud Shape Completion [completion; Github; NeurIPS]
  • Category-Level 6D Object Pose Estimation in the Wild: A Semi-Supervised Learning Approach and A New Dataset [pose estimation; PyTorch; NeurIPS]
  • GenSDF: Two-Stage Learning of Generalizable Signed Distance Functions [representation; PyTorch; NeurIPS]
  • Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training [self-supervised; PyTorch; NeurIPS]
  • Learning Inter-Superpoint Affinity for Weakly Supervised 3D Instance Segmentation [seg; PyTorch; ACCV]
  • Point Cloud Upsampling via Cascaded Refinement Network [upsampling; Github; ACCV]
  • Learning Object-level Point Augmentor for Semi-supervised 3D Object Detection [det; PyTorch; BMVC]
  • APSNet: Attention Based Point Cloud Sampling [sampling; PyTorch; BMVC]
  • Data Augmentation-free Unsupervised Learning for 3D Point Cloud Understanding [unsupervised; PyTorch; BMVC]
  • COARSE3D: Class-Prototypes for Contrastive Learning in Weakly-Supervised 3D Point Cloud Segmentation [seg; PyTorch; BMVC]
  • GraphReg: Dynamical Point Cloud Registration with Geometry-aware Graph Signal Processing [registration; Github; TIP]
  • 3D Cascade RCNN: High Quality Object Detection in Point Clouds [det; PyTorch; TIP]
  • Real-time 3D Single Object Tracking with Transformer [tracking; PyTorch; TMM]
  • Push-the-Boundary: Boundary-aware Feature Propagation for Semantic Segmentation of 3D Point Clouds [seg; Github; 3DV]
  • SRFeat: Learning Locally Accurate and Globally Consistent Non-Rigid Shape Correspondence [non-rigid, matching; Github; 3DV]
  • SVNet: Where SO(3) Equivariance Meets Binarization on Point Cloud Representation [cls, seg; Github; 3DV]
  • Bidirectional Feature Globalization for Few-shot Semantic Segmentation of 3D Point Cloud Scenes [seg; 3DV]
  • Arbitrary Point Cloud Upsampling with Spherical Mixture of Gaussians [upsampling; 3DV]
  • CC-3DT: Panoramic 3D Object Tracking via Cross-Camera Fusion [tracking; Project; CoRL]
  • Towards Long-Tailed 3D Detection [det; CoRL]
  • CoBEVT: Cooperative Bird's Eye View Semantic Segmentation with Sparse Transformers [seg; PyTorch; CoRL]
  • SSL-Lanes: Self-Supervised Learning for Motion Forecasting in Autonomous Driving [motion forecasting; PyTorch; CoRL]
  • MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance Segmentation [seg; ICME]
  • CENet: Toward Concise and Efficient LiDAR Semantic Segmentation for Autonomous Driving [seg; PyTorch; ICME]
  • Learning to Generate 3D Shapes from a Single Example [generation; PyTorch; SIGGRAPH Aisa]
  • Shape Completion with Points in the Shadow [completion; SIGGRAPH Aisa]
  • ImLoveNet: Misaligned Image-supported Registration Network for Low-overlap Point Cloud Pairs [registration; SIGGRAPH]
  • Graph-DETR3D: Rethinking Overlapping Regions for Multi-View 3D Object Detection [det; ACM MM]
  • GRASP-Net: Geometric Residual Analysis and Synthesis for Point Cloud Compression [compressing; Github; ACM MMW]
  • Paint and Distill: Boosting 3D Object Detection with Semantic Passing Network [det; Github; ACM MM]
  • You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptors [registration; PyTorch; ACM MM]
  • Boosting Single-Frame 3D Object Detection by Simulating Multi-Frame Point Clouds [det; ACM MM]
  • LiSnowNet: Real-time Snow Removal for LiDAR Point Cloud [autonomous driving; IROS]
  • Patchwork++: Fast and Robust Ground Segmentation Solving Partial Under-Segmentation Using 3D Point Cloud [seg; Github; IROS]
  • Efficient Spatial-Temporal Information Fusion for LiDAR-Based 3D Moving Object Segmentation [seg; PyTorch; IROS]
  • 3D Part Assembly Generation with Instance Encoded Transformer [part assembly; IROS]
  • BoxGraph: Semantic Place Recognition and Pose Estimation from 3D LiDAR [place recognition; IROS]
  • Benchmarking and Analyzing Point Cloud Classification under Corruptions [cls; PyTorch; ICML]
  • AutoAlign: Pixel-Instance Feature Aggregation for Multi-Modal 3D Object Detection [det; IJCAI]
  • BiCo-Net: Regress Globally, Match Locally for Robust 6D Pose Estimation [pose estimation; IJCAI]
  • Spatiality-guided Transformer for 3D Dense Captioning on Point Clouds [captioning; Github; IJCAI]
  • Robust Point Cloud Segmentation with Noisy Annotations [seg; PyTorch; TPAMI]
  • Robust Point Cloud Registration Framework Based on Deep Graph Matching [registration; PyTorch; TPAMI]
  • TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Driving [autonomous driving; PyTorch; TPAMI]
  • RIGA: Rotation-Invariant and Globally-Aware Descriptors for Point Cloud Registration [registration; TPAMI]
  • Detecting Rotated Objects as Gaussian Distributions and Its 3-D Generalization [det; TPAMI]
  • Graph Neural Network and Spatiotemporal Transformer Attention for 3D Video Object Detection from Point Clouds [det; TPAMI]
  • Ret3D: Rethinking Object Relations for Efficient 3D Object Detection in Driving Scenes [det; TPAMI]
  • PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution [seg, det; TPAMI]
  • Pixel2Mesh++: 3D Mesh Generation and Refinement from Multi-View Images [generation, mesh; TPAMI]
  • Refine-Net: Normal Refinement Neural Network for Noisy Point Clouds [normal refinement; PyTorch; TPAMI]
  • PMP-Net++: Point Cloud Completion by Transformer-Enhanced Multi-step Point Moving Paths [completion; PyTorch; TPAMI]
  • A New Outlier Removal Strategy Based on Reliability of Correspondence Graph for Fast Point Cloud Registration [registration; Github; TPAMI]
  • DSGN++: Exploiting Visual-Spatial Relation for Stereo-based 3D Detectors [det; PyTorch; TPAMI]
  • Snowflake Point Deconvolution for Point Cloud Completion and Generation with Skip-Transformer [completion; PyTorch; TPAMI]
  • WSDesc: Weakly Supervised 3D Local Descriptor Learning for Point Cloud Registration [registration; PyTorch; TVCG]
  • Point Set Self-Embedding [embedding; Github; TVCG]
  • SoftPool++: An Encoder-Decoder Network for Point Cloud Completion [completion; IJCV]
  • RIConv++: Effective Rotation Invariant Convolutions for 3D Point Clouds Deep Learning [cls, seg, retrieval; IJCV]
  • Enhanced Prototypical Learning for Unsupervised Domain Adaptation in LiDAR Semantic Segmentation [seg; ICRA]
  • Multi-Class 3D Object Detection with Single-Class Supervision [det; ICRA]
  • Learning 6-DoF Object Poses to Grasp Category-level Objects by Language Instructions [pose estimation; Project; ICRA]
  • HiTPR: Hierarchical Transformer for Place Recognition in Point Cloud [place recognition; ICRA]
  • RMS-FlowNet: Efficient and Robust Multi-Scale Scene Flow Estimation for Large-Scale Point Clouds [scene flow; ICRA]
  • Variable Rate Compression for Raw 3D Point Clouds [compression; Github; ICRA]
  • Hindsight is 2020: Leveraging Past Traversals to Aid 3D Perception [det; PyTorch; ICLR]
  • WeakM3D: Towards Weakly Supervised Monocular 3D Object Detection [det, monocular; Github; ICLR]
  • Partial Wasserstein Adversarial Network for Non-rigid Point Set Registration [non-rigid, registration; ICLR]
  • A Conditional Point Diffusion-Refinement Paradigm for 3D Point Cloud Completion [completion; PyTorch; ICLR]
  • Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Framework [cls, seg; PyTorch; ICLR]
  • MonoDistill: Learning Spatial Features for Monocular 3D Object Detection [det, monocular; Github; ICLR]
  • Robust 3D Object Detection in Cold Weather Conditions [det; IV]
  • urban_road_filter: A real-time LIDAR-based urban road and sidewalk detection algorithm for autonomous vehicles [det, autonomous driving; Github; Video; Sensors]
  • GraffMatch: Global Matching of 3D Lines and Planes for Wide Baseline LiDAR Registration [registration; RAL]
  • Exploiting More Information in Sparse Point Cloud for 3D Single Object Tracking [tracking; Github; RAL]
  • DeepFusionMOT: A 3D Multi-Object Tracking Framework Based on Camera-LiDAR Fusion with Deep Association [tracking; Github; RAL]
  • BIMS-PU: Bi-Directional and Multi-Scale Point Cloud Upsampling [upsampling; RAL]
  • Receding Moving Object Segmentation in 3D LiDAR Data Using Sparse 4D Convolutions [seg; Github; RAL]
  • Temporal Point Cloud Completion with Pose Disturbance [completion; RAL]
  • Semantic Segmentation-Assisted Instance Feature Fusion for Multi-Level 3D Part Instance Segmentation [seg; Tensorflow; CVM]
  • Semi-supervised 3D shape segmentation with multilevel consistency and part substitution [seg; Tensorflow; CVM]
  • Point cloud completion on structured feature map with feedback network [completion; CVM]
  • TorchSparse: Efficient Point Cloud Inference Engine [engine; PyTorch; MLSys]
• arXiv
  • TiG-BEV: Multi-view BEV 3D Object Detection via Target Inner-Geometry Learning [det; Github]
  • MM-3DScene: 3D Scene Understanding by Customizing Masked Modeling with Informative-Preserved Reconstruction and Self-Distilled Consistency [seg, det]
  • ADAS: A Simple Active-and-Adaptive Baseline for Cross-Domain 3D Semantic Segmentation [seg; PyTorch]
  • OBMO: One Bounding Box Multiple Objects for Monocular 3D Object Detection [det; Github]
  • 3D Point Cloud Pre-training with Knowledge Distillation from 2D Images [pre-training]
  • Point-E: A System for Generating 3D Point Clouds from Complex Prompts [generation; PyTorch]
  • DETR4D: Direct Multi-View 3D Object Detection with Sparse Attention [det]
  • MAELi -- Masked Autoencoder for Large-Scale LiDAR Point Clouds [self-supervised]
  • ULIP: Learning Unified Representation of Language, Image and Point Cloud for 3D Understanding [cls; Github]
  • BEV-MAE: Bird's Eye View Masked Autoencoders for Outdoor Point Cloud Pre-training [pre-training]
  • Synthetic-to-Real Domain Generalized Semantic Segmentation for 3D Indoor Point Clouds [seg]
  • SemanticBEVFusion: Rethink LiDAR-Camera Fusion in Unified Bird's-Eye View Representation for 3D Object Detection [det]
  • SDFusion: Multimodal 3D Shape Completion, Reconstruction, and Generation [completion, reconstruction, generation; Github]
  • Fast Point Cloud Generation with Straight Flows [generation]
  • PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained Image-Language Models [seg]
  • GeoUDF: Surface Reconstruction from 3D Point Clouds via Geometry-guided Distance Representation [reconstruction; PyTorch]
  • Language-driven Open-Vocabulary 3D Scene Understanding [open-vocabulary; Github]
  • 3D Point Positional Encoding for Multi-Camera 3D Object Detection Transformers [det]
  • Meta Architecure for Point Cloud Analysis [seg, cls; PyTorch]
  • Learning 3D Scene Priors with 2D Supervision [layout, shape; Project]
  • PVT3D: Point Voxel Transformers for Place Recognition from Sparse Lidar Scans [place recognition]
  • ONeRF: Unsupervised 3D Object Segmentation from Multiple Views [seg, reconstruction]
  • AeDet: Azimuth-invariant Multi-view 3D Object Detection [det]
  • Sparse4D: Multi-view 3D Object Detection with Sparse Spatial-Temporal Fusion [det]
  • MatrixVT: Efficient Multi-Camera to BEV Transformation for 3D Perception [seg, det]
  • PointCLIP V2: Adapting CLIP for Powerful 3D Open-world Learning [pre-training; Github]
  • Adaptive Edge-to-Edge Interaction Learning for Point Cloud Analysis [cls, seg]
  • 3D-QueryIS: A Query-based Framework for 3D Instance Segmentation [seg]
  • You Only Label Once: 3D Box Adaptation from Point Cloud to Image via Semi-Supervised Learning [det]
  • Towards 3D Object Detection with 2D Supervision [det]
  • Structured Knowledge Distillation Towards Efficient and Compact Multi-View 3D Detection [det]
  • Point-DAE: Denoising Autoencoders for Self-supervised Point Cloud Learning [self-supervised; Github]
  • 3D Reconstruction of Multiple Objects by mmWave Radar on UAV [reconstruction]
  • Ground Plane Matters: Picking Up Ground Plane Prior in Monocular 3D Object Detection [det]
  • Point-Voxel Adaptive Feature Abstraction for Robust Point Cloud Classification [cls; PyTorch]
  • LidarAugment: Searching for Scalable 3D LiDAR Data Augmentations [augmentation]
  • Domain Adaptation in 3D Object Detection with Gradual Batch Alternation Training [det]
  • Zero-shot Point Cloud Segmentation by Transferring Geometric Primitives [seg]
  • Common Corruption Robustness of Point Cloud Detectors: Benchmark and Enhancement [det]
  • Mask3D for 3D Semantic Instance Segmentation [seg]
  • STD: Stable Triangle Descriptor for 3D place recognition [place recognition; Github]
  • Neural Wavelet-domain Diffusion for 3D Shape Generation [generation]
  • FusionRCNN: LiDAR-Camera Fusion for Two-stage 3D Object Detection [det; Github]
  • LidarMultiNet: Towards a Unified Multi-task Network for LiDAR Perception [det]
  • 4DenoiseNet: Adverse Weather Denoising from Adjacent Point Clouds [autonomous driving]
  • Point Cloud Registration-Driven Robust Feature Matching for 3D Siamese Object Tracking [tracking]
  • CRAFT: Camera-Radar 3D Object Detection with Spatio-Contextual Fusion Transformer [det]
  • M^2-3DLaneNet: Multi-Modal 3D Lane Detection [det]
  • Multi-modal Streaming 3D Object Detection [det]
  • ISS: Image as Stetting Stone for Text-Guided 3D Shape Generation [generation]
  • CAMO-MOT: Combined Appearance-Motion Optimization for 3D Multi-Object Tracking with Camera-LiDAR Fusion [tracking]
  • SEFormer: Structure Embedding Transformer for 3D Object Detection [det]
  • Benchmarking Robustness of 3D Point Cloud Recognition Against Common Corruptions [robustness; PyTorch]
  • Scatter Points in Space: 3D Detection from Multi-view Monocular Images [det]
  • Towards Accurate Reconstruction of 3D Scene Shape from A Single Monocular Image [reconstruction; PyTorch]
  • PV-RCNN++: Semantical Point-Voxel Feature Interaction for 3D Object Detection [det]
  • Voxurf: Voxel-based Efficient and Accurate Neural Surface Reconstruction [reconstruction]
  • Pix4Point: Image Pretrained Transformers for 3D Point Cloud Understanding [pretraining; Github]
  • Bridging the View Disparity of Radar and Camera Features for Multi-modal Fusion 3D Object Detection [det]
  • AGO-Net: Association-Guided 3D Point Cloud Object Detection Network [det]
  • PeRFception: Perception using Radiance Fields [NeRF; PyTorch]
  • Quality Matters: Embracing Quality Clues for Robust 3D Multi-Object Tracking [tracking]
  • STS: Surround-view Temporal Stereo for Multi-view 3D Detection [det]
  • PointDP: Diffusion-driven Purification against Adversarial Attacks on 3D Point Cloud Recognition [adversarial attack]
  • PersDet: Monocular 3D Detection in Perspective Bird’s-Eye-View [det]
  • InterTrack: Interaction Transformer for 3D Multi-Object Tracking [tracking]
  • An Empirical Study of Pseudo-Labeling for Image-based 3D Object Detection [det]
  • RWSeg: Cross-graph Competing Random Walks for Weakly Supervised 3D Instance Segmentation [seg]
  • Exploring Point-BEV Fusion for 3D Point Cloud Object Tracking with Transformer [tracking; PyTorch]
  • RadSegNet: A Reliable Approach to Radar Camera Fusion [autonomous driving]
  • Aerial Monocular 3D Object Detection [det; Project]
  • TransPillars: Coarse-to-Fine Aggregation for Multi-Frame 3D Object Detection [det]
  • Point-McBert: A Multi-choice Self-supervised Framework for Point Cloud Pre-training [self-supervised]
  • MV-FCOS3D++: Multi-View Camera-Only 4D Object Detection with Pretrained Monocular Backbones [det; PyTorch]
  • DBQ-SSD: Dynamic Ball Query for Efficient 3D Object Detection [det]
  • On the Robustness of 3D Object Detectors [det]
  • Omni3D: A Large Benchmark and Model for 3D Object Detection in the Wild [det; PyTorch]
  • Boosting 3D Object Detection via Object-Focused Image Fusion [det; PyTorch]
  • Det6D: A Ground-Aware Full-Pose 3D Object Detector for Improving Terrain Robustness [det; Github]
  • Consistency of Implicit and Explicit Features Matters for Monocular 3D Object Detection [det, monocular]
  • UniFormer: Unified Multi-view Fusion Transformer for Spatial-Temporal Representation in Bird's-Eye-View [autonomous driving]
  • PointNorm: Normalization is All You Need for Point Cloud Analysis [cls, seg; Github]
  • SpOT: Spatiotemporal Modeling for 3D Object Tracking [tracking]
  • CP3: Unifying Point Cloud Completion by Pretrain-Prompt-Predict Paradigm [completion]
  • Implicit Autoencoder for Point Cloud Self-supervised Representation Learning [self-supervised; PyTorch]
  • Learning to Register Unbalanced Point Pairs [registration]
  • Learning Spatial and Temporal Variations for 4D Point Cloud Segmentation [seg]
  • MT-Net Submission to the Waymo 3D Detection Leaderboard [det]
  • Masked Surfel Prediction for Self-Supervised Point Cloud Learning [self-supervised; Github]
  • GLENet: Boosting 3D Object Detectors with Generative Label Uncertainty Estimation [det]
  • Open-Vocabulary 3D Detection via Image-level Class and Debiased Cross-modal Contrastive Learning [det]
  • ORA3D: Overlap Region Aware Multi-view 3D Object Detection [det]
  • Masked Autoencoders in 3D Point Cloud Representation Learning [self-supervised]
  • Masked Autoencoders for Self-Supervised Learning on Automotive Point Clouds [self-supervised]
  • SARNet: Semantic Augmented Registration of Large-Scale Urban Point Clouds [registration; PyTorch]
  • HM3D-ABO: A Photo-realistic Dataset for Object-centric Multi-view 3D Reconstruction [reconstruction; Github]
  • NeuRIS: Neural Reconstruction of Indoor Scenes Using Normal Priors [reconstruction; Project]
  • Unseen Object 6D Pose Estimation: A Benchmark and Baselines [pose estimation; Project]
  • LidarMutliNet: Unifying LiDAR Semantic Segmentation, 3D Object Detection, and Panoptic Segmentation in a Single Multi-task Network [seg, det]
  • Polar Parametrization for Vision-based Surround-View 3D Detection [det; Github]
  • Reconstruct from Top View: A 3D Lane Detection Approach based on Geometry Structure Prior [autonomous driving]
  • Voxel-MAE: Masked Autoencoders for Pre-training Large-scale Point Clouds [det; Github]
  • VectorMapNet: End-to-end Vectorized HD Map Learning [autonomous driving; Github]
  • A Simple Baseline for BEV Perception Without LiDAR [autonomous driving; PyTorch]
  • Level 2 Autonomous Driving on a Single Device: Diving into the Devils of Openpilot [autonomous driving; Github]
  • LET-3D-AP: Longitudinal Error Tolerant 3D Average Precision for Camera-Only 3D Detection [det]
  • Semi-signed neural fitting for surface reconstruction from unoriented point clouds [reconstruction]
  • PatchComplete: Learning Multi-Resolution Patch Priors for 3D Shape Completion on Unseen Categories [completion; Project]
  • AGConv: Adaptive Graph Convolution on 3D Point Clouds [cls, seg; PyTorch]
  • VN-Transformer: Rotation-Equivariant Attention for Vector Neurons [cls, motion forecasting]
  • Depth Estimation Matters Most: Improving Per-Object Depth Estimation for Monocular 3D Detection and Tracking [det, tracking]
  • SHRED: 3D Shape Region Decomposition with Learned Local Operations [decomposition]
  • Fast and Robust Non-Rigid Registration Using Accelerated Majorization-Minimization [registration, non-rigid; Github]
  • PETRv2: A Unified Framework for 3D Perception from Multi-Camera Images [autonomous driving]
  • SparseDet: Towards End-to-End 3D Object Detection [det]
  • Benchmarking the Robustness of LiDAR-Camera Fusion for 3D Object Detection [det; Github]
  • Towards Efficient 3D Object Detection with Knowledge Distillation [det]
  • OpenCalib: A multi-sensor calibration toolbox for autonomous driving [calibration, autonomous driving; Github]
  • BEVFusion: A Simple and Robust LiDAR-Camera Fusion Framework [det; PyTorch]
  • BEVerse: Unified Perception and Prediction in Birds-Eye-View for Vision-Centric Autonomous Driving [autonomous driving; Github]
  • Continual learning on 3D point clouds with random compressed rehearsal [continual learning]
  • Panoptic-PHNet: Towards Real-Time and High-Precision LiDAR Panoptic Segmentation via Clustering Pseudo Heatmap [seg]
  • Weakly Supervised 3D Point Cloud Segmentation via Multi-Prototype Learning [seg]
  • Cost-Aware Comparison of LiDAR-based 3D Object Detectors [det]
  • Open-Set Semi-Supervised Learning for 3D Point Cloud Understanding [cls, seg]
  • APP-Net: Auxiliary-point-based Push and Pull Operations for Efficient Point Cloud Classification [cls]
  • Unsupervised Domain Adaptation for Monocular 3D Object Detection via Self-Training [det, monocular]
  • CPGNet: Cascade Point-Grid Fusion Network for Real-Time LiDAR Semantic Segmentation [seg]
  • RangeUDF: Semantic Surface Reconstruction from 3D Point Clouds [reconstruction; Github]
  • Dynamic Point Cloud Denoising via Gradient Fields [denoising]
  • Stress-Testing LiDAR Registration [registration; Github]
  • M^2BEV: Multi-Camera Joint 3D Detection and Segmentation with Unified Birds-Eye View Representation [det, seg; Project]
  • POS-BERT: Point Cloud One-Stage BERT Pre-Training [cls, seg; Github]
  • DODA: Data-oriented Sim-to-Real Domain Adaptation for 3D Indoor Semantic Segmentation [seg]
  • ImpDet: Exploring Implicit Fields for 3D Object Detection [det]
  • Learning a Structured Latent Space for Unsupervised Point Cloud Completion [completion]
  • Self-Supervised Point Cloud Representation Learning with Occlusion Auto-Encoder [self-supervised; Github]
  • Towards 3D Scene Understanding by Referring Synthetic Models [transfer learning]
  • Unsupervised Learning of 3D Semantic Keypoints with Mutual Reconstruction [keypoints]
  • Self-supervised Point Cloud Completion on Real Traffic Scenes via Scene-concerned Bottom-up Mechanism [completion]
  • FUTR3D: A Unified Sensor Fusion Framework for 3D Detection [det; PyTorch]
  • 3DAC: Learning Attribute Compression for Point Clouds [compression]
  • CodedVTR: Codebook-based Sparse Voxel Transformer with Geometric Guidance [seg]
  • DetMatch: Two Teachers are Better Than One for Joint 2D and 3D Semi-Supervised Object Detection [det; Github]
  • PointAttN: You Only Need Attention for Point Cloud Completion [completion; PyTorch]
  • Deep learning for radar data exploitation of autonomous vehicle [radar, autonomous vehicle]
  • LiDAR-based 4D Panoptic Segmentation via Dynamic Shifting Network [seg]
  • CVFNet: Real-time 3D Object Detection by Learning Cross View Features [det]
  • PD-Flow: A Point Cloud Denoising Framework with Normalizing Flows [denoising]
  • An Empirical Investigation of 3D Anomaly Detection and Segmentation [anomaly detection; PyTorch]
  • A Lightweight and Detector-free 3D Single Object Tracker on Point Clouds [tracking]
  • DisARM: Displacement Aware Relation Module for 3D Detection [det]
  • Dense Voxel Fusion for 3D Object Detection [det]
  • Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptors [anomaly detection]
  • PointMatch: A Consistency Training Framework for Weakly Supervised Semantic Segmentation of 3D Point Clouds [seg]
  • LighTN: Light-weight Transformer Network for Performance-overhead Tradeoff in Point Cloud Downsampling [downsampling]
  • Distillation with Contrast is All You Need for Self-Supervised Point Cloud Representation Learning [self-supervised]
  • Edge-Selective Feature Weaving for Point Cloud Matching [correspondence; PyTorch-lightning]
  • Neighborhood-aware Geometric Encoding Network for Point Cloud Registration [registration; PyTorch]
  • Boosting Monocular Depth Estimation with Sparse Guided Points [monocular, depth estimation; Github]
  • Trajectory Forecasting from Detection with Uncertainty-Aware Motion Encoding [autonomous platforms]
  • TPC: Transformation-Specific Smoothing for Point Cloud Models [attack]
  • Self-supervised Point Cloud Registration with Deep Versatile Descriptors [registration]
  • CP-Net: Contour-Perturbed Reconstruction Network for Self-Supervised Point Cloud Learning [self-supervised]
  • Box2Seg: Learning Semantics of 3D Point Clouds with Box-Level Supervision [seg]

2021

• ICCV
  • erception-Aware Multi-Sensor Fusion for 3D LiDAR Semantic Segmentation [seg]
  • MVTN: Multi-View Transformation Network for 3D Shape Recognition [cls, retrieval; PyTorch]
  • Self-Supervised Pretraining of 3D Features on any Point-Cloud [self-supervised; PyTorch]
  • MGNet: Monocular Geometric Scene Understanding for Autonomous Driving [autonomous driving; PyTorch]
  • FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detection [monocular, det; mmdet3d]
  • Progressive Seed Generation Auto-encoder for Unsupervised Point Cloud Learning [unsupervised]
  • Guided Point Contrastive Learning for Semi-supervised Point Cloud Semantic Segmentation [seg]
  • Pyramid Point Cloud Transformer for Large-Scale Place Recognition [place recognition; Github]
  • Distinctiveness oriented Positional Equilibrium for Point Cloud Registration [registration]
  • Feature Interactive Representation for Point Cloud Registration [registration]
  • DeepPRO: Deep Partial Point Cloud Registration of Objects [registration]
  • LSG-CPD: Coherent Point Drift with Local Surface Geometry for Point Cloud Registration [registration; matlab]
  • Provably Approximated Point Cloud Registration [registration]
  • Point Transformer [seg, cls; PyTorch-unofficial]
  • Point Cloud Augmentation with Weighted Local Transformations [augmentation; PyTorch]
  • PCAM: Product of Cross-Attention Matrices for Rigid Registration of Point Clouds [registration; PyTorch]
  • An End-to-End Transformer Model for 3D Object Detection [det; PyTorch]
  • Sampling Network Guided Cross-Entropy Method for Unsupervised Point Cloud Registration [registration; Github]
  • Deep Hough Voting for Robust Global Registration [registration; PyTorch]
  • Pyramid R-CNN: Towards Better Performance and Adaptability for 3D Object Detection [det]
  • Voxel Transformer for 3D Object Detection [det]
  • Learning Inner-Group Relations on Point Clouds [cls, seg]
  • Spatio-temporal Self-Supervised Representation Learning for 3D Point Clouds [self-supervised; Github]
  • 4D-Net for Learned Multi-Modal Alignment [det]
  • AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection [monocular, det; Github]
  • A Robust Loss for Point Cloud Registration [registration]
  • OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud Registration [registration]
  • Improving 3D Object Detection with Channel-wise Transformer [det; Github]
  • Voxel-based Network for Shape Completion by Leveraging Edge Generation [completion; Github]
  • Exploring Simple 3D Multi-Object Tracking for Autonomous Driving [tracking]
  • ME-PCN: Point Completion Conditioned on Mask Emptiness [completion]
  • Deep Hybrid Self-Prior for Full 3D Mesh Generation [generation]
  • Fine-grained Semantics-aware Representation Enhancement for Self-supervised Monocular Depth Estimation [monocular, depth; Github]
  • StructDepth: Leveraging the structural regularities for self-supervised indoor depth estimation [monocular, depth; PyTorch]
  • Vis2Mesh: Efficient Mesh Reconstruction from Unstructured Point Clouds of Large Scenes with Learned Virtual View Visibility [reconstruction; Github]
  • PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformers [completion; PyTorch]
  • Adaptive Graph Convolution for Point Cloud Analysis [cls, seg; PyTorch]
  • RandomRooms: Unsupervised Pre-training from Synthetic Shapes and Randomized Layouts for 3D Object Detection [det]
  • Self-supervised Monocular Depth Estimation for All Day Images using Domain Separation [monocular, depth; Github]
  • Instance Segmentation in 3D Scenes using Semantic Superpoint Tree Networks [seg; Github]
  • Is Pseudo-Lidar needed for Monocular 3D Object detection? [monocular, det; Github]
  • Towards Efficient Point Cloud Graph Neural Networks Through Architectural Simplification
  • Walk in the Cloud: Learning Curves for Point Clouds Shape Analysis [cls, seg; Github]
  • AdaFit: Rethinking Learning-based Normal Estimation on Point Clouds [normal estimation; Github]
  • Fog Simulation on Real LiDAR Point Clouds for 3D Object Detection in Adverse Weather [det; Github]
  • Box-Aware Feature Enhancement for Single Object Tracking on Point Clouds [tracking; Github]
  • SnowflakeNet: Point Cloud Completion by Snowflake Point Deconvolution with Skip-Transformer [completion; Github]
  • DRINet: A Dual-Representation Iterative Learning Network for Point Cloud Segmentation [seg]
  • RangeDet: In Defense of Range View for LiDAR-based 3D Object Detection [det; Github]
  • Hierarchical Aggregation for 3D Instance Segmentation [seg; Github]
  • Sparse-to-dense Feature Matching: Intra and Inter domain Cross-modal Learning in Domain Adaptation for 3D Semantic Segmentation [seg; Github]
  • Group-Free 3D Object Detection via Transformers [det; PyTorch]
  • VMNet: Voxel-Mesh Network for Geodesic-Aware 3D Semantic Segmentation [seg; Github]
  • Learning with Noisy Labels for Robust Point Cloud Segmentation [seg; Github]
  • Geometry Uncertainty Projection Network for Monocular 3D Object Detection [det, monocular]
  • ReDAL: Region-based and Diversity-aware Active Learning for Point Cloud Semantic Segmentation [seg; PyTorch]
  • Unsupervised Domain Adaptive 3D Detection with Multi-Level Consistency [det; OpenPCDet]
  • Unsupervised Point Cloud Pre-Training via View-Point Occlusion, Completion [pre-training; Github]
  • HRegNet: A Hierarchical Network for Large-scale Outdoor LiDAR Point Cloud Registration [registration; PyTorch]
  • Score-Based Point Cloud Denoising [denoising]
  • Probabilistic Monocular 3D Human Pose Estimation with Normalizing Flows [monocular, pose; Github]
  • A Technical Survey and Evaluation of Traditional Point Cloud Clustering Methods for LiDAR Panoptic Segmentation [seg; Github]
  • The Devil is in the Task: Exploiting Reciprocal Appearance-Localization Features for Monocular 3D Object Detection [monocular, det]
• CVPR
  • Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling [self-supervised; PyTorch]
  • Self-Supervised Learning on 3D Point Clouds by Learning Discrete Generative Models [self-supervised]
  • Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation [seg; PyTorch]
  • PointAugmenting: Cross-Modal Augmentation for 3D Object Detection [det]
  • PVGNet: A Bottom-Up One-Stage 3D Object Detector with Integrated Multi-Level Features [det]
  • MetaSets: Meta-Learning on Point Sets for Generalizable Representations [domain]
  • LiDAR-based Panoptic Segmentation via Dynamic Shifting Network [seg; PyTorch]
  • PMP-Net: Point Cloud Completion by Learning Multi-step Point Moving Paths [completion; PyTorch]
  • CorrNet3D: Unsupervised End-to-end Learning of Dense Correspondence for 3D Point Clouds [correspondence; PyTorch-lightning]
  • StickyPillars: Robust and Efficient Feature Matching on Point Clouds using Graph Neural Networks [registration]
  • To the Point: Efficient 3D Object Detection in the Range Image with Graph Convolution Kernels [det]
  • RSN: Range Sparse Net for Efficient, Accurate LiDAR 3D Object Detection [det]
  • Point Cloud Upsampling via Disentangled Refinement [upsampling; Github]
  • Omni-supervised Point Cloud Segmentation via Gradual Receptive Field Component Reasoning [seg]
  • Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contexts [seg; PyTorch]
  • PU-GCN: Point Cloud Upsampling using Graph Convolutional Networks [upsampling; Tensorflow]
  • Self-Point-Flow: Self-Supervised Scene Flow Estimation from Points Clouds with Optimal Transport and Random Walk [scene flow]
  • SAIL-VOS 3D: A Synthetic Dataset and Baselines for Object Detection and 3D Mesh Reconstruction from Video Data [reconstruction]
  • HCRF-Flow: Scene Flow from Point Clouds with Continuous High-order CRFs and Position-aware Flow Embedding [scene flow]
  • 3D Spatial Recognition without Spatially Labeled 3D [det, seg]
  • LASR: Learning Articulated Shape Reconstruction from a Monocular Video [reconstruction, monocular]
  • VoxelContext-Net: An Octree based Framework for Point Cloud Compression [compression]
  • Unsupervised 3D Shape Completion through GAN Inversion [completion; PyTorch]
  • KeypointDeformer: Unsupervised 3D Keypoint Discovery for Shape Control [Github]
  • Multi-Modal Fusion Transformer for End-to-End Autonomous Driving [autonomous driving; PyTorch]
  • Self-Supervised Pillar Motion Learning for Autonomous Driving [autonomous driving; Github]
  • Variational Relational Point Completion Network [completion; PyTorch]
  • Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds [det; Github]
  • RPSRNet: End-to-End Trainable Rigid Point Set Registration Network using Barnes-Hut 2D-Tree Representation [registration]
  • Objects are Different: Flexible Monocular 3D Object Detection [det; Github]
  • FESTA: Flow Estimation via Spatial-Temporal Attention for Scene Point Clouds [scene flow]
  • HVPR: Hybrid Voxel-Point Representation for Single-stage 3D Object Detection [det]
  • Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation [seg; PyTorch]
  • ReAgent: Point Cloud Registration using Imitation and Reinforcement Learning [registration; PyTorch]
  • LiDAR R-CNN: An Efficient and Universal 3D Object Detector [det; Github]
  • Equivariant Point Network for 3D Point Cloud Analysis [Github]
  • PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Clouds [cls, det; Github]
  • Depth-conditioned Dynamic Message Propagation for Monocular 3D Object Detection [det; Github]
  • Delving into Localization Errors for Monocular 3D Object Detection [det; Github]
  • M3DSSD: Monocular 3D Single Stage Object Detector [det; Github]
  • Cycle4Completion: Unpaired Point Cloud Completion using Cycle Transformation with Missing Region Coding [completion]
  • Monte Carlo Scene Search for 3D Scene Understanding
  • Semantic Segmentation for Real Point Cloud Scenes via Bilateral Augmentation and Adaptive Fusion [seg; Github]
  • PointDSC: Robust Point Cloud Registration using Deep Spatial Consistency [registration; PyTorch]
  • ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detection [det; OpenPCDet]
  • Robust Point Cloud Registration Framework Based on Deep Graph Matching [registration; Github]
  • RfD-Net: Point Scene Understanding by Semantic Instance Reconstruction [reconstruction]
  • MultiBodySync: Multi-Body Segmentation and Motion Estimation via 3D Scan Synchronization [motion analysis; Github]
  • TPCN: Temporal Point Cloud Networks for Motion Forecasting [motion forecasting]
  • Self-supervised Geometric Perception [self-supervised; Github]
  • PointGuard: Provably Robust 3D Point Cloud Classification [cls]
  • Point 4D Transformer Networks for Spatio-Temporal Modeling in Point Cloud Videos
  • SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud [det; Github]
  • Center-based 3D Object Detection and Tracking [det, tracking; PyTorch]
  • 3DIoUMatch: Leveraging IoU Prediction for Semi-Supervised 3D Object Detection [det; PyTorch]
  • Style-based Point Generator with Adversarial Rendering for Point Cloud Completion [completion]
  • FFB6D: A Full Flow Bidirectional Fusion Network for 6D Pose Estimation [pose estimation; Github]
  • Diffusion Probabilistic Models for 3D Point Cloud Generation [generation; Github]
  • GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D Object Pose Estimation [pose estimation; Github]
  • PCLs: Geometry-aware Neural Reconstruction of 3D Pose with Perspective Crop Layers [reconstruction]
  • PREDATOR: Registration of 3D Point Clouds with Low Overlap [registration; PyTorch]
  • SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registration [registration; Github]
  • Categorical Depth Distribution Network for Monocular 3D Object Detection [det]
  • Multimodal Motion Prediction with Stacked Transformers [motion prediction; Github]
  • GrooMeD-NMS: Grouped Mathematically Differentiable NMS for Monocular 3D Object Detection [det; PyTorch]
  • Model-based 3D Hand Reconstruction via Self-Supervised Learning [reconstruction]
  • MonoRUn: Monocular 3D Object Detection by Self-Supervised Reconstruction and Uncertainty Propagation [det; Github]
  • Deep Implicit Moving Least-Squares Functions for 3D Reconstruction [reconstruction; Tensorflow]
  • Skeleton Merger: an Unsupervised Aligned Keypoint Detector [keypoint; PyTorch]
  • Single Image Depth Prediction with Wavelet Decomposition [depth; PyTorch]
  • 3D Shape Generation with Grid-based Implicit Functions [generation]
• Others
  • PRA-Net: Point Relation-Aware Network for 3D Point Cloud Analysis [cls, seg; PyTorch; TIP]
  • DeepUME: Learning the Universal Manifold Embedding for Robust Point Cloud Registration [registration; PyTorch; BMVC]
  • On Automatic Data Augmentation for 3D Point Cloud Classification [augmentation, cls; BMVC]
  • Self-Supervised Point Cloud Completion via Inpainting [completion; BMVC]
  • Two Heads are Better than One: Geometric-Latent Attention for Point Cloud Classification and Segmentation [cls, seg; BMVC]
  • 3D Object Tracking with Transformer [tracking; Github; BMVC]
  • Cascading Feature Extraction for Fast Point Cloud Registration [registration; BMVC]
  • PolarStream: Streaming Lidar Object Detection and Segmentation with Polar Pillars [det, seg; PyTorch; NeurIPS]
  • Revisiting 3D Object Detection From an Egocentric Perspective [det; NeurIPS]
  • Density-aware Chamfer Distance as a Comprehensive Metric for Point Cloud Completion [completion; PyTorch; NeurIPS]
  • Multimodal Virtual Point 3D Detection [det; PyTorch; NeurIPS]
  • 3D Siamese Voxel-to-BEV Tracker for Sparse Point Clouds [tracking; Github; NeurIPS]
  • Voxel-based 3D Detection and Reconstruction of Multiple Objects from a Single Image [monocular, det, reconstruction; NeurIPS]
  • Accurate Point Cloud Registration with Robust Optimal Transport [registration; Github; NeurIPS]
  • CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registration [registration; PyTorch; NeurIPS]
  • Object DGCNN: 3D Object Detection using Dynamic Graphs [det; Github; NeurIPS]
  • Multi-modal Trajectory Prediction for Autonomous Driving with Semantic Map and Dynamic Graph Attention Network [autonomous driving; NeurIPS]
  • Probabilistic and Geometric Depth: Detecting Objects in Perspective [det; mmdet3d; CoRL]
  • DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries [det; Github; CoRL]
  • Self-supervised Point Cloud Prediction Using 3D Spatio-temporal Convolutional Networks [autonomous driving; Github; CoRL]
  • Semi-supervised 3D Object Detection via Temporal Graph Neural Networks [det]
  • GASCN: Graph Attention Shape Completion Network [completion; 3DV]
  • DurLAR: A High-Fidelity 128-Channel LiDAR Dataset with Panoramic Ambient and Reflectivity Imagery for Multi-Modal Autonomous Driving Applications [monocular, autonomous driving; Github; 3DV]
  • Learning Iterative Robust Transformation Synchronization [transformation synchronization; Github; 3DV]
  • DPC: Unsupervised Deep Point Correspondence via Cross and Self Construction [correspondence; PyTorch; 3DV]
  • DeepBBS: Deep Best Buddies for Point Cloud Registration [registration; PyTorch; 3DV]
  • Similarity-Aware Fusion Network for 3D Semantic Segmentation [seg; Github; IROS]
  • Cross-modality Discrepant Interaction Network for RGB-D Salient Object Detection [det; ACM MM]
  • From Voxel to Point: IoU-guided 3D Object Detection for Point Cloud with Voxel-to-Point Decoder [det; Github; ACM MM]
  • Anchor-free 3D Single Stage Detector with Mask-Guided Attention for Point Cloud [det; Github; ACM MM]
  • Hierarchical View Predictor: Unsupervised 3D Global Feature Learning through Hierarchical Prediction among Unordered Views [unsupervised; ACM MM]
  • SSPU-Net: Self-Supervised Point Cloud Upsampling via Differentiable Rendering [upsampling; Github; ACM MM]
  • Self-Contrastive Learning with Hard Negative Sampling for Self-supervised Point Cloud Learning [self-supervised; ACM MM]
  • Neighbor-Vote: Improving Monocular 3D Object Detection through Neighbor Distance Voting [monocular, det; ACM MM]
  • Fast and Robust Registration of Partially Overlapping Point Clouds [registration; PyTorch; RAL]
  • Graph-Guided Deformation for Point Cloud Completion [completion; RAL]
  • GIDSeg: Learning 3D Segmentation from Sparse Annotations via Hierarchical Descriptors [seg; RAL]
  • Planning with Learned Dynamic Model for Unsupervised Point Cloud Registration [registration; IJCAI]
  • PointLIE: Locally Invertible Embedding for Point Cloud Sampling and Recovery [sampling; Github; IJCAI]
  • Unsupervised Shape Completion via Deep Prior in the Neural Tangent Kernel Perspective [completion; TOG]
  • Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR-based Perception [seg, det; PyTorch; TPAMI]
  • Point Cloud Instance Segmentation with Semi-supervised Bounding-Box Mining [seg; TPAMI]
  • Learning Semantic Segmentation of Large-Scale Point Clouds with Random Sampling [seg; TPAMI]
  • Fast and Robust Iterative Closest Point [registration; Github; TPAMI]
  • MonoGRNet: A General Framework for Monocular 3D Object Detection [monocular, det; TPAMI]
  • PSTNet: Point Spatio-Temporal Convolution on Point Cloud Sequences [action recognition, seg; ICLR]
  • PC2WF: 3D Wireframe Reconstruction from Raw Point Clouds [wireframe; ICLR]
  • Self-Guided Instance-Aware Network for Depth Completion and Enhancement [depth; ICRA]
  • FGR: Frustum-Aware Geometric Reasoning for Weakly Supervised 3D Vehicle Detection [det; Github; ICRA]
  • Exploiting Local Geometry for Feature and Graph Construction for Better 3D Point Cloud Processing with Graph Neural Networks [cls, seg; ICRA]
  • 3D3L: Deep Learned 3D Keypoint Detection and Description for LiDARs [keypoint; Github; ICRA]
  • NDT-Transformer: Large-Scale 3D Point Cloud Localisation using the Normal Distribution Transform Representation [localisation; ICRA]
  • Volumetric Propagation Network: Stereo-LiDAR Fusion for Long-Range Depth Estimation [depth estimation; ICRA]
  • YOLOStereo3D: A Step Back to 2D for Efficient Stereo 3D Detection [det; PyTorch; ICRA]
  • ERASOR: Egocentric Ratio of Pseudo Occupancy-based Dynamic Object Removal for Static 3D Point Cloud Map Building [static map; ICRA]
  • CloudAAE: Learning 6D Object Pose Regression with On-line Data Synthesis on Point Clouds [pose estimation; Tensorflow; ICRA]
  • Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline [cls; PyTorch; ICML]
  • PointCutMix: Regularization Strategy for Point Cloud Classification [cls; code; ICML]
  • Investigate Indistinguishable Points in Semantic Segmentation of 3D Point Cloud [seg; Github; AAAI]
  • PointINet: Point Cloud Frame Interpolation Network [frame interpolation; PyTorch; AAAI]
  • Boundary-Aware Geometric Encoding for Semantic Segmentation of Point Clouds [seg; code; AAAI]
  • Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection [det; AAAI]
  • Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point Cloud [cls, seg; AAAI]
  • CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloud [det; PyTorch; AAAI]
  • Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completion [seg; Github; AAAI]
  • labelCloud: A Lightweight Domain-Independent Labeling Tool for 3D Object Detection in Point Clouds [labeling tool; CAD]
  • CenterFusion: Center-based Radar and Camera Fusion for 3D Object Detection [det; PyTorch; WACV]
• arXiv
  • COTReg: Coupled Optimal Transport based Point Cloud Registration [registration]
  • iSeg3D: An Interactive 3D Shape Segmentation Tool [seg, annotation tool]
  • Multi-View Partial (MVP) Point Cloud Challenge 2021 on Completion and Registration: Methods and Results [completion, registration; PyTorch]
  • BEVDet: High-Performance Multi-Camera 3D Object Detection in Bird-Eye-View [det; Github]
  • Revisiting Transformation Invariant Geometric Deep Learning: Are Initial Representations All You Need? [transformation invariant]
  • High-Fidelity Point Cloud Completion with Low-Resolution Recovery and Noise-Aware Upsampling [completion]
  • EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detection [det]
  • Domain Adaptation on Point Clouds via Geometry-Aware Implicits [domain adaptation]
  • Full Transformer Framework for Robust Point Cloud Registration with Deep Information Interaction [registration; Github]
  • Immortal Tracker: Tracklet Never Dies [tracking; Github]
  • Robust Partial-to-Partial Point Cloud Registration in a Full Range [registration; PyTorch]
  • Semi-supervised Implicit Scene Completion from Sparse LiDAR [completion; PyTorch]
  • Multi-instance Point Cloud Registration by Efficient Correspondence Clustering [registration]
  • Multiway Non-rigid Point Cloud Registration via Learned Functional Map Synchronization [registration, non-rigid; Github]
  • PU-Transformer: Point Cloud Upsampling Transformer [upsampling]
  • GenReg: Deep Generative Method for Fast Point Cloud Registration [registration]
  • Deep Point Cloud Reconstruction [reconstruction]
  • MFM-Net: Unpaired Shape Completion Network with Multi-stage Feature Matching [completion]
  • Sparse Tensor-based Multiscale Representation for Point Cloud Geometry Compression [compression; Github]
  • What Stops Learning-based 3D Registration from Working in the Real World? [registration]
  • CpT: Convolutional Point Transformer for 3D Point Cloud Processing [cls, seg]
  • RAANet: Range-Aware Attention Network for LiDAR-based 3D Object Detection with Auxiliary Density Level Estimation [det; PyTorch]
  • IMFNet: Interpretable Multimodal Fusion for Point Cloud Registration [registration]
  • SimpleTrack: Understanding and Rethinking 3D Multi-object Tracking [track; Github]
  • DRINet++: Efficient Voxel-as-point Point Cloud Segmentation [seg]
  • Robust 3D Scene Segmentation through Hierarchical and Learnable Part-Fusion [seg]
  • DFC: Deep Feature Consistency for Robust Point Cloud Registration [registration]
  • Interpreting Representation Quality of DNNs for 3D Point Cloud Processing
  • CPSeg: Cluster-free Panoptic Segmentation of 3D LiDAR Point Clouds [seg]
  • Structure Information is the Key: Self-Attention RoI Feature Extractor in 3D Object Detection [det; Github]
  • Deep Models with Fusion Strategies for MVP Point Cloud Registration [registration]
  • Improved Pillar with Fine-grained Feature for 3D Object Detection [det]
  • 3D Object Detection Combining Semantic and Geometric Features from Point Clouds [det]
  • How to Build a Curb Dataset with LiDAR Data for Autonomous Driving [autonomous driving]
  • 3D-FCT: Simultaneous 3D Object Detection and Tracking Using Feature Correlation [det, tracking]
  • PQ-Transformer: Jointly Parsing 3D Objects and Layouts from Point Clouds [PyTorch]
  • Differentiable Convolution Search for Point Cloud Processing [cls, seg]
  • SMAC-Seg: LiDAR Panoptic Segmentation via Sparse Multi-directional Attention Clustering [seg]
  • GP-S3Net: Graph-based Panoptic Sparse Semantic Segmentation Network [seg]
  • SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generation [det]
  • Progressive Coordinate Transforms for Monocular 3D Object Detection [monocular, det; Github]
  • UPDesc: Unsupervised Point Descriptor Learning for Robust Registration [registration]
  • Investigating Attention Mechanism in 3D Point Cloud Object Detection [det; Github]
  • Real-Time Anchor-Free Single-Stage 3D Detection with IoU-Awareness [det]
  • Learning Geometry-Guided Depth via Projective Modeling for Monocular 3D Object Detection [det; Github]
  • CarveNet: Carving Point-Block for Complex 3D Shape Completion [completion]
  • CKConv: Learning Feature Voxelization for Point Cloud Analysis [cls, seg]
  • DV-Det: Efficient 3D Point Cloud Object Detection with Dynamic Voxelization [det]
  • Unsupervised Domain Adaptation in LiDAR Semantic Segmentation with Self-Supervision and Gated Adapters [seg]
  • Dynamic Convolution for 3D Point Cloud Instance Segmentation[seg; PyTorch]
  • Beyond Farthest Point Sampling in Point-Wise Analysis [sampling]
  • Learn to Learn Metric Space for Few-Shot Segmentation of 3D Shapes [seg]
  • Multi-Modality Task Cascade for 3D Object Detection [det; Github]
  • Point Cloud Registration using Representative Overlapping Points [registration; PyTorch]
  • “Zero Shot” Point Cloud Upsampling [upsampling]
  • Shape registration in the time of transformers [registration]
  • 3D Shape Registration Using Spectral Graph Embedding and Probabilistic Matching [registration]
  • Image2Point: 3D Point-Cloud Understanding withPretrained 2D ConvNets [cls, seg; Github]
  • Z2P: Instant Rendering of Point Clouds [rendering]
  • TransLoc3D : Point Cloud based Large-scale Place Recognition using Adaptive Receptive Fields [place recognition]
  • Generalisable and distinctive 3D local deep descriptors for point cloud registration [registration]
  • Deep Weighted Consensus (DWC) Dense correspondence confidence maps for 3D shape registration [registration]
  • M3DETR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers [det]
  • Boundary-Aware 3D Object Detection from Point Clouds [det]
  • Geometry-aware data augmentation for monocular 3D object detection [det]
  • OCM3D: Object-Centric Monocular 3D Object Detection [det]
  • Towards Efficient Graph Convolutional Networks for Point Cloud Handling [network; Github]
  • Occlusion Guided Self-supervised Scene Flow Estimation on 3D Point Clouds [scene flow]
  • A Learnable Self-supervised Task for Unsupervised Domain Adaptation on Point Clouds [UDA]
  • View-Guided Point Cloud Completion [completion]
  • One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation [seg]
  • Potential Convolution: Embedding Point Clouds into Potential Fields [cls, seg]
  • 3D-MAN: 3D Multi-frame Attention Network for Object Detection [det]
  • SIENet: Spatial Information Enhancement Network for 3D Object Detection from Point Cloud [det; Github]
  • 3D Point Cloud Registration with Multi-Scale Architecture and Self-supervised Fine-tuning [registration; Github]
  • Multi-view 3D Reconstruction with Transformer [reconstruction]
  • X-view: Non-egocentric Multi-View 3D Object Detector [det]
  • RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR Point Cloud Segmentation [seg]
  • 3DMNDT: 3D multi-view registration method based on the normal distributions transform [registration]
  • SparsePoint: Fully End-to-End Sparse 3D Object Detector [det]
  • S3Net: 3D LiDAR Sparse Semantic Segmentation Network [seg]
  • Lite-HDSeg: LiDAR Semantic Segmentation Using Lite Harmonic Dense Convolutions [seg]
  • R-PointHop: A Green, Accurate and Unsupervised Point Cloud Registration Method [registration; Github]
  • Model-free Vehicle Tracking and State Estimation in Point Cloud Sequences [autonomous driving; Github]
  • MapFusion: A General Framework for 3D Object Detection with HDMaps [det]
  • Offboard 3D Object Detection from Point Cloud Sequences [det]
  • A Simple and Efficient Multi-task Network for 3D Object Detection and Road Understanding [det; PyTorch]
  • IRON: Invariant-based Highly Robust Point Cloud Registration [registration]
  • EllipsoidNet: Ellipsoid Representation for Point Cloud Classification and Segmentation [cls, seg]
  • Pseudo-labeling for Scalable 3D Object Detection [det]
  • LiDAR-based Recurrent 3D Semantic Segmentation with Temporal Memory Alignment [seg]
  • Scalable Scene Flow from Point Clouds in the Real World [scene flow]
  • InstanceRefer: Cooperative Holistic Understanding for Visual Grounding on Point Clouds through Instance Multi-level Contextual Referring [visual grounding]
  • FPS-Net: A Convolutional Fusion Network for Large-Scale LiDAR Point Cloud Segmentation [seg]
  • P2-Net: Joint Description and Detection of Local Features for Pixel and Point Matching [matching]
  • UnsupervisedR&R: Unsupervised Point Cloud Registration via Differentiable Rendering [registration; PyTorch]
  • Attention Models for Point Clouds in Deep Learning: A Survey [attention]
  • EfficientLPS: Efficient LiDAR Panoptic Segmentation [seg]
  • HyperPocket: Generative Point Cloud Completion [completion]
  • Point-set Distances for Learning Representations of 3D Point Clouds [representation]
  • DPointNet: A Density-Oriented PointNet for 3D Object Detection in Point Clouds [det]
  • PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection [det; OpenPCDet]
  • Self-Attention Based Context-Aware 3D Object Detection [det; PyTorch]
  • A two-stage data association approach for 3D Multi-object Tracking [tracking]
  • The Devils in the Point Clouds: Studying the Robustness of Point Cloud Convolutions [seg]
  • Joint Learning of 3D Shape Retrieval and Deformation
  • Efficient 3D Point Cloud Feature Learning for Large-Scale Place Recognition [place recognition; Tensorflow]

2020

• ECCV
  • Object as Hotspots: An Anchor-Free 3D Object Detection Approach via Firing of Hotspots [det]
  • PointMixup: Augmentation for point cloud [augmentation, cls; PyTorch]
  • Deformable PV-RCNN: Improving 3D Object Detection with Learned Deformations [det; PyTorch]
  • Unsupervised Learning of Category-Specific Symmetric 3D Keypoints from Point Sets [keypoints]
  • Weakly-supervised 3D Shape Completion in the Wild [completion]
  • SoftPoolNet: Shape Descriptor for Point Cloud Completion and Classification [completion, cls; Github]
  • Detail Preserved Point Cloud Completion via Separated Feature Aggregation [completion; Tensorflow]
  • PointPWC-Net: A Coarse-to-Fine Network for Supervised and Self-Supervised Scene Flow Estimation on 3D Point Clouds [flow estimation; PyTorch]
  • JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds [seg; Tensorflow]
  • A Closer Look at Local Aggregation Operators in Point Cloud Analysis [cls, seg; Code]
  • Instance-Aware Embedding for Point Cloud Instance Segmentation [seg]
  • Multimodal Shape Completion via Conditional Generative Adversarial Networks [completion; PyTorch]
  • GRNet: Gridding Residual Network for Dense Point Cloud Completion [completion; PyTorch]
  • 3D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View Spatial Feature Fusion for 3D Object Detection [det]
  • SSN: Shape Signature Networks for Multi-class Object Detection from Point Clouds [det; Github]
  • Pillar-based Object Detection for Autonomous Driving [det, autonomous driving; Tensorflow]
  • EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection [det; PyTorch]
  • Finding Your (3D) Center: 3D Object Detection Using a Learned Loss [det; Tensorflow]
  • Weakly Supervised 3D Object Detection from Lidar Point Cloud [det; PyTorch]
  • H3DNet: 3D Object Detection Using Hybrid Geometric Primitives [det; Tensorflow]
  • Generative Sparse Detection Networks for 3D Single-shot Object Detection [det; Github]
  • Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution [seg, det; PyTorch]
  • DeepGMR: Learning Latent Gaussian Mixture Models for Registration [registration; PyTorch]
  • Quaternion Equivariant Capsule Networks for 3D Point Clouds [PyTorch]
  • PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding [unsupervised; cls, seg, det; PyTorch]
  • Convolutional Occupancy Networks [reconstruction; PyTorch]
  • Iterative Distance-Aware Similarity Matrix Convolution with Mutual-Supervised Point Elimination for Efficient Point Cloud Registration [registration; PyTorch]
  • Progressive Point Cloud Deconvolution Generation Network [generation; github]
  • Reinforced Axial Refinement Network for Monocular 3D Object Detection [det, monocular]
  • Monocular 3D Object Detection via Feature Domain Adaptation [det, monocular]
  • Improving 3D Object Detection through Progressive Population Based Augmentation [det]
  • An LSTM Approach to Temporal 3D Object Detection in LiDAR Point Clouds [det]
  • Rotation-robust Intersection over Union for 3D Object Detection
  • DPDist: Comparing Point Clouds Using Deep Point Cloud Distance
• CVPR
  • End-to-end pseudo-lidar for image-based 3d object detection [det; PyTorch]
  • PointPainting: Sequential Fusion for 3D Object Detection [det]
  • 3DSSD: Point-based 3D Single Stage Object Detector [det; Tensorflow]
  • A Hierarchical Graph Network for 3D Object Detection on Point Clouds [det]
  • Deep Geometric Functional Maps: Robust Feature Learning for Shape Correspondence [correspondences; Tensorflow]
  • Deep Global Registration [registration; PyTorch]
  • 3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation [seg; Github]
  • PointGMM: a Neural GMM Network for Point Clouds [generation, registration; PyTorch]
  • Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud [det; Tensorflow]
  • ImVoteNet: Boosting 3D Object Detection in Point Clouds with Image Votes [det]
  • OccuSeg: Occupancy-aware 3D Instance Segmentation [seg]
  • Fusion-Aware Point Convolution for Online Semantic 3D Scene Segmentation [seg; PyTorch]
  • MLCVNet: Multi-Level Context VoteNet for 3D Object Detection [det; PyTorch]
  • Going Deeper with Lean Point Networks [seg; PyTorch]
  • Point Cloud Completion by Skip-attention Network with Hierarchical Folding [completion]
  • Unsupervised Learning of Intrinsic Structural Representation Points [PyTorch]
  • PF-Net: Point Fractal Network for 3D Point Cloud Completion [completion; PyTorch]
  • PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection [det; code]
  • Adaptive Hierarchical Down-Sampling for Point Cloud Classification [downsampling, cls]
  • SA-SSD: Structure Aware Single-stage 3D Object Detection from Point Cloud [det; PyTorch]
  • 3DRegNet: A Deep Neural Network for 3D Point Registration [registration; Tensorflow]
  • MINA: Convex Mixed-Integer Programming for Non-Rigid Shape Alignment [non-rigid alignment]
  • SampleNet: Differentiable Point Cloud Sampling [sample, cls, registration, reconstruction; PyTorch]
  • Learning multiview 3D point cloud registration [multiview registration; PyTorch]
  • Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondences [registration; PyTorch]
  • PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling [cls, seg; Tensorflow]
  • Global-Local Bidirectional Reasoning for Unsupervised Representation Learning of 3D Point Clouds [unsupervised; cls; PyTorch]
  • Grid-GCN for Fast and Scalable Point Cloud Learning [cls, seg; mxnet]
  • FPConv: Learning Local Flattening for Point Convolution [cls, seg; PyTorch]
  • PointAugment: an Auto-Augmentation Framework for Point Cloud Classification [cls, retrieval; github]
  • RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds [seg; Tensorflow]
  • Weakly Supervised Semantic Point Cloud Segmentation:Towards 10X Fewer Labels [weakly supervised; seg; Tensorflow]
  • PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation [seg; PyTorch]
  • Learning to Segment 3D Point Clouds in 2D Image Space [seg; Keras]
  • PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation [seg; PyTorch]
  • D3Feat: Joint Learning of Dense Detection and Description of 3D Local Features [keypoints, registration; Tensorflow, PyTorch]
  • RPM-Net: Robust Point Matching using Learned Features [registration; PyTorch]
  • Cascaded Refinement Network for Point Cloud Completion [completion; Tensorflow]
  • P2B: Point-to-Box Network for 3D Object Tracking in Point Clouds [tracking; PyTorch]
  • An Efficient PointLSTM for Point Clouds Based Gesture Recognition [gesture; PyTorch]
• Others
  • Group Contextual Encoding for 3D Point Clouds [det, cls; PyTorch; NeurIPS]
  • CaSPR: Learning Canonical Spatiotemporal Point Cloud Representations [dynamic sequences; Github; NeurIPS]
  • Skeleton-bridged Point Completion: From Global Inference to Local Adjustment [completion; NeurIPS]
  • Self-Supervised Few-Shot Learning on Point Clouds [cls, seg; NeurIPS]
  • Rotation-Invariant Local-to-Global Representation Learning for 3D Point Cloud [cls; NeurIPS]
  • PIE-NET: Parametric Inference of Point Cloud Edges [edge det; NeurIPS]
  • Spherical Kernel for Efficient Graph Convolution on 3D Point Clouds [cls, seg; Tensorflow; TPAMI]
  • From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network [det; PyTorch; TPAMI]
  • Unpaired Point Cloud Completion on Real Scans using Adversarial Training [completion; Tensorflow; ICLR]
  • AdvectiveNet: An Eulerian-Lagrangian Fluidic Reservoir for Point Cloud Processing [cls, seg; PyTorch; ICLR]
  • Tranquil Clouds: Neural Networks for Learning Temporally Coherent Features in Point Clouds [ICLR]
  • PI-RCNN: An Efficient Multi-Sensor 3D Object Detector with Point-Based Attentive Cont-Conv Fusion Module [det; AAAI]
  • MSN: Morphing and Sampling Network for Dense Point Cloud Completion [completion; PyTorch; AAAI]
  • TANet: Robust 3D Object Detection from Point Clouds with Triple Attention [det; PyTorch; AAAI]
  • JSNet: Joint Instance and Semantic Segmentation of 3D Point Clouds [seg; Tensorflow]
  • Point2Node: Correlation Learning of Dynamic-Node for Point Cloud Feature Modeling [cls, seg; AAAI]
  • Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution [cls, seg, matching; AAAI]
  • Differentiable Manifold Reconstruction for Point Cloud Denoising [denoising; PyTorch; ACM MM]
  • Weakly Supervised 3D Object Detection from Point Clouds [det; Tensorflow; ACM MM]
  • Unsupervised Detection of Distinctive Regions on 3D Shapes [unsupervised; Tensorflow; TOG]
  • SegVoxelNet: Exploring Semantic Context and Depth-aware Features for 3D Vehicle Detection from Point Cloud [det; ICRA]
  • Dilated Point Convolutions: On the Receptive Field Size of Point Convolutions on 3D Point Clouds [seg, cls; Project; ICRA]
  • Semantic Graph Based Place Recognition for 3D Point Clouds [place recognition; PyTorch; IROS]
  • End-to-End 3D Point Cloud Learning for Registration Task Using Virtual Correspondences [registration; PyTorch; IROS]
  • Correspondence Matrices are Underrated [registration, correspondence; PyTorch; 3DV]
  • Learning Rotation-Invariant Representations of Point Clouds Using Aligned Edge Convolutional Neural Networks [cls, seg; 3DV]
  • PanoNet3D: Combining Semantic and Geometric Understanding for LiDAR Point Cloud Detection [det; 3DV]
  • FKAConv: Feature-Kernel Alignment for Point Cloud Convolution [conv, cls, seg; PyTorch; ACCV]
  • Sparse Convolutions on Continuous Domains for Point Cloud and Event Stream Networks [conv, cls; ACCV]
  • Reconstructing Human Body Mesh from Point Clouds by Adversarial GP Network [reconstruction; ACCV]
  • Bi-Directional Attention for Joint Instance and Semantic Segmentation in Point Clouds [seg; Tensorflow; ACCV]
  • SDP-Net: Scene Flow Based Real-time Object Detection and Prediction from Sequential 3D Point Clouds [det; ACCV]
  • Best Buddies Registration for Point Clouds [registration; PyTorch; ACCV]
  • HPGCNN: Hierarchical Parallel Group Convolutional Neural Networks for Point Clouds Processing [conv, cls, seg; ACCV]
  • SAUM: Symmetry-Aware Upsampling Module for Consistent Point Cloud Completion [completion; Tensorflow; ACCV]
  • Fast and Automatic Registration of Terrestrial Point Clouds Using 2D Line Features [registration; Remote Sensing]
  • ConvPoint: Continuous Convolutions for Point Cloud Processing [cls, seg; PyTorch; Computers & Graphics]
• arXiv
  • Multi-Modality Cut and Paste for 3D Object Detection [det; PyTorch]
  • Self-Supervised Learning for Domain Adaptation on Point Clouds [cls, seg]
  • SALA: Soft Assignment Local Aggregation for 3D Semantic Segmentation [seg]
  • Compositional Prototype Network with Multi-view Comparision for Few-Shot Point Cloud Semantic Segmentation [seg]
  • Geometric robust descriptor for 3D point cloud [registration, cls, seg]
  • PCT: Point Cloud Transformer [cls, seg, normal estimation; Jittor]
  • Point Transformer(Nico) [cls, seg]
  • Deterministic PointNetLK for Generalized Registration [registration]
  • OcCo: Pre-Training by Completing Point Clouds [pre-training, completion; Github]
  • Learning 3D-3D Correspondences for One-shot Partial-to-partial Registration [registration]
  • PRE-TRAINING BY COMPLETING POINT CLOUDS [pre-training, cls, seg; Github]
  • Continuous Geodesic Convolutions for Learning on 3D Shapes [descriptor, match, seg]
  • Multi-Resolution Graph Neural Network for Large-Scale Pointcloud Segmentation [seg]
  • A Density-Aware PointRCNN for 3D Objection Detection in Point Clouds [det]
  • TEASER: Fast and Certifiable Point Cloud Registration [registration; Github]
  • Part-Aware Data Augmentation for 3D Object Detection in Point Cloud [det, augmentation; PyTorch]

2019

• ICCV
  • M3D-RPN: Monocular 3D Region Proposal Network for Object Detection [det]
  • Disentangling Monocular 3D Object Detection [det, monocular]
  • Total Denoising: Unsupervised Learning of 3D Point Cloud Cleaning [denoising; Tensorflow]
  • 3D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutions [generation; PyTorch]
  • STD: Sparse-to-Dense 3D Object Detector for Point Cloud [det]
  • USIP: Unsupervised Stable Interest Point Detection from 3D Point Clouds [keypoints, registration; PyTorch]
  • LPD-Net: 3D Point Cloud Learning for Large-Scale Place Recognition and Environment Analysis [place recognition]
  • Unsupervised Multi-Task Feature Learning on Point Clouds [cls, seg]
  • Multi-Angle Point Cloud-VAE: Unsupervised Feature Learning for 3D Point Clouds from Multiple Angles by Joint Self-Reconstruction and Half-to-Half Prediction [unsupervised, cls, generation, seg, completion]
  • SemanticKITTI: A Dataset for Semantic Scene Understanding of LiDAR Sequences [dataset]
  • MeteorNet: Deep Learning on Dynamic 3D Point Cloud Sequences [cls, seg, flow estimation; Tensorflow]
  • DeepGCNs: Can GCNs Go as Deep as CNNs? [seg; Tensorflow]
  • VV-NET: Voxel VAE Net with Group Convolutions for Point Cloud Segmentation [seg; Github]
  • Interpolated Convolutional Networks for 3D Point Cloud Understanding [cls, seg]
  • Dynamic Points Agglomeration for Hierarchical Point Sets Learning [cls, seg]
  • ShellNet: Efficient Point Cloud Convolutional Neural Networks using Concentric Shells Statistics [cls, seg; Tensorflow]
  • Fast Point R-CNN [det]
  • Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data [dataset; cls; Tensorflow]
  • KPConv: Flexible and Deformable Convolution for Point Clouds [cls, seg; code]
  • Fully Convolutional Geometric Features [match; PyTorch]
  • Deep Closest Point: Learning Representations for Point Cloud Registration [registration; PyTorch]
  • DeepICP: An End-to-End Deep Neural Network for 3D Point Cloud Registration [registration]
  • Efficient and Robust Registration on the 3D Special Euclidean Group [registration]
  • Hierarchical Point-Edge Interaction Network for Point Cloud Semantic Segmentation [seg]
  • DensePoint: Learning Densely Contextual Representation for Efficient Point Cloud Processing [cls, retrieval, seg, normal estimation; PyTorch]
  • DUP-Net: Denoiser and Upsampler Network for 3D Adversarial Point Clouds Defense [cls]
  • Efficient Learning on Point Clouds with Basis Point Sets [cls, registration; PyTorch]
  • PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows [generation, reconstruction; Pytorch
  • PU-GAN: a Point Cloud Upsampling Adversarial Network [upsampling, reconstruction; Project]
  • 3D Point Cloud Learning for Large-scale Environment Analysis and Place Recognition [retrieval, place recognition]
  • Deep Hough Voting for 3D Object Detection in Point Clouds [det; PyTorch]
  • Exploring the Limitations of Behavior Cloning for Autonomous Driving [autonomous driving; Pytorch]
• CVPR
  • Deep Fitting Degree Scoring Network for Monocular 3D Object Detection [det, monocular]
  • Multi-Task Multi-Sensor Fusion for 3D Object Detection [det]
  • LiDAR-based Online 3D Video Object Detection with Graph-based Message Passing and Spatiotemporal Transformer Attention [det; Github]
  • TopNet: Structural Point Cloud Decoder [completion; Github]
  • FlowNet3D: Learning Scene Flow in 3D Point Clouds [scene flow; Tensorflow]
  • Occupancy Networks: Learning 3D Reconstruction in Function Space [reconstruction]
  • Associatively Segmenting Instances and Semantics in Point Clouds [seg; Tensorflow]
  • 3D Point Capsule Networks [autoencoder; PyTorch]
  • Patch-based Progressive 3D Point Set Upsampling [upsampling; Tensorflow, PyTorch]
  • Generating 3D Adversarial Point Clouds [adversary; Tensorflow]
  • RL-GAN-Net: A Reinforcement Learning Agent Controlled GAN Network for Real-Time Point Cloud Shape Completion [completion; PyTorch]
  • GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in Point Cloud [seg; Tensorflow]
  • JSIS3D: Joint Semantic-Instance Segmentation of 3D Point Clouds with Multi-Task Pointwise Networks and Multi-Value Conditional Random Fields [seg; PyTorch]
  • 3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans [seg; PyTorch]
  • Learning Transformation Synchronization [transformation synchronization, registration; PyTorch]
  • SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondences [registration; Github]
  • Learning Transformation Synchronization [reconstruction; PyTorch]
  • 3D Local Features for Direct Pairwise Registration [registration]
  • DeepMapping: Unsupervised Map Estimation From Multiple Point Clouds [registration; Github]
  • Relation-Shape Convolutional Neural Network for Point Cloud Analysis [cls, seg, normal estimation; PyTorch]
  • Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNN [cls, det; Tensorflow]
  • 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks [seg; PyTorch]
  • PCAN: 3D Attention Map Learning Using Contextual Information for Point Cloud Based Retrieval [retrieval; Tensorflow]
  • Attentional PointNet for 3D-Object Detection in Point Clouds [det; PyTorch]
  • Octree guided CNN with Spherical Kernels for 3D Point Clouds [cls, seg; Github]
  • A-CNN: Annularly Convolutional Neural Networks on Point Clouds [cls, seg; Tensorflow]
  • ClusterNet: Deep Hierarchical Cluster Network with Rigorously Rotation-Invariant Representation for Point Cloud Analysis [cls]
  • Graph Attention Convolution for Point Cloud Semantic Segmentation [seg; PyTorch-unofficial]
  • PointWeb: Enhancing Local Neighborhood Features for Point Cloud Processing [seg, cls; PyTorch]
  • Modeling Point Clouds with Self-Attention and Gumbel Subset Sampling [cls, seg, gesture]
  • Learning to Sample [sample, cls, retrieval, reconstruction; Tensorflow]
  • PointConv: Deep Convolutional Networks on 3D Point Clouds [cls, seg; Tensorflow]
  • The Perfect Match: 3D Point Cloud Matching With Smoothed Densities [match; code]
  • PointNetLK: Point Cloud Registration using PointNet [registration; PyTorch]
  • PointRCNN: 3D Object Proposal Generation and Detection From Point Cloud [det; PyTorch]
  • PointPillars: Fast Encoders for Object Detection From Point Clouds [det; Pytorch]
  • Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving [depth estimation, det; github]
  • ApolloCar3D: A Large 3D Car Instance Understanding Benchmark for Autonomous Driving [dataset, autonomous driving]
  • Stereo R-CNN based 3D Object Detection for Autonomous Driving [det, autonomous driving; github]
  • Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction [det, autonomous driving; Tesorflow]
  • LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving [det]
  • GS3D: An Efficient 3D Object Detection Framework for Autonomous Driving [det, autonomous driving]
  • L3-Net: Towards Learning based LiDAR Localization for Autonomous Driving [autonomous driving]
  • Iterative Transformer Network for 3D Point Cloud [pose, cls, seg; Tensorflow]
• Others
  • End-to-End Multi-View Fusion for 3D Object Detection in LiDAR Point Clouds [det; CoRL]
  • PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud Representation [domain adaptation; PyTorch; NeurIPS]
  • Learning elementary structures for 3D shape generation and matching [generation, matching; NeurIPS]
  • Self-Supervised Deep Learning on Point Clouds by Reconstructing Space [self-supervised, cls, seg; NeurIPS]
  • Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds [seg; Tensorflow; NeurIPS]
  • PRNet: Self-Supervised Learning for Partial-to-Partial Registration [registration, cls; PyTorch; NeurIPS]
  • Point-Voxel CNN for Efficient 3D Deep Learning [seg, det; PyTorch; NeurIPS]
  • L2G Auto-encoder: Understanding Point Clouds by Local-to-Global Reconstruction with Hierarchical Self-Attention [autoencoder; ACM MM]
  • Deep Cascade Generation on Point Sets [generation; PyTorch; IJCAI]
  • A Polynomial-time Solution for Robust Registration with Extreme Outlier Rates [registration; RSS]
  • Dynamic Graph CNN for Learning on Point Clouds [cls, seg; Github; TOG]
  • SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud [seg; Tensorflow; ICRA]
  • Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection [det; PyTorch; IROS]
  • RangeNet++: Fast and Accurate LiDAR Semantic Segmentation [seg; PyTorch; IROS]
  • IoU Loss for 2D/3D Object Detection [det; 3DV]
  • AlignNet-3D: Fast Point Cloud Registration of Partially Observed Objects [registration; Tensorflow; 3DV]
  • Dense 3D Point Cloud Reconstruction Using a Deep Pyramid Network [reconstruction; WACV]
• arXiv
  • Class-balanced Grouping and Sampling for Point Cloud 3D Object Detection [det]
  • PCRNet: Point Cloud Registration Network using PointNet Encoding [registration; PyTorch, Tensorflow]
  • LSANet: Feature Learning on Point Sets by Local Spatial Aware Layer [cls, seg; Tensorflow]
  • Adversarial Sensor Attack on LiDAR-based Perception in Autonomous Driving [autonomous driving]
  • Linked Dynamic Graph CNN: Learning on Point Cloud via Linking Hierarchical Features [cls, seg; Tensorflow]

2018

• CVPR
  • PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation [det]
  • Learning 3D Shape Completion From Laser Scan Data With Weak Supervision [completion; Github]
  • Deep Parametric Continuous Convolutional Neural Networks [seg, motion estimation(lidar flow)]
  • Attentional ShapeContextNet for Point Cloud Recognition [cls, seg]
  • A Papier-Mâché Approach to Learning 3D Surface Generation [generation; PyTorch]
  • Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs [seg; PyTorch]
  • FoldingNet: Point Cloud Auto-encoder via Deep Grid Deformation [autoencoder, unsupervised; code]
  • FeaStNet: Feature-Steered Graph Convolutions for 3D Shape Analysis [correspondence, seg; Tensorflow]
  • PointNetVLAD: Deep Point Cloud Based Retrieval for Large-Scale Place Recognition [retrieval, place recognition; Tensorflow]
  • PU-Net: Point Cloud Upsampling Network [upsampling; Tensorflow]
  • SGPN: Similarity Group Proposal Network for 3D Point Cloud Instance Segmentation [seg; Tensorflow]
  • Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling [cls, seg; code]
  • Tangent Convolutions for Dense Prediction in 3D [seg; Tensorflow]
  • PointGrid: A Deep Network for 3D Shape Understanding [cls, seg; Tensorflow]
  • 3D Semantic Segmentation with Submanifold Sparse Convolutional Networks [seg; Github]
  • Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs [seg; PyTorch]
  • SPLATNet: Sparse Lattice Networks for Point Cloud Processing [seg; Caffe]
  • Pointwise Convolutional Neural Networks [cls, seg; Tensorflow]
  • SO-Net: Self-Organizing Network for Point Cloud Analysis [autoencoder, cls, seg; PyTorch]
  • Recurrent Slice Networks for 3D Segmentation of Point Clouds [seg; PyTorch]
  • PPFNet: Global Context Aware Local Features for Robust 3D Point Matching [registration]
  • PIXOR: Real-Time 3D Object Detection From Point Clouds [det; PyTorch]
  • Frustum PointNets for 3D Object Detection From RGB-D Data [det; Tensorflow]
  • VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection [det]
  • 3D-RCNN: Instance-Level 3D Object Reconstruction via Render-and-Compare [reconstruction]
  • Multi-Level Fusion Based 3D Object Detection From Monocular Images [det]
• ECCV
  • Complex-YOLO: Real-time 3D Object Detection on Point Clouds [det; PyTorch; ECCVW]
  • 3D-CODED : 3D Correspondences by Deep Deformation [matching; PyTorch]
  • SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters [cls, seg; Tensorflow]
  • 3DContextNet: K-d Tree Guided Hierarchical Learning of Point Clouds Using Local and Global Contextual Cues [seg, cls]
  • Multiresolution Tree Networks for 3D Point Cloud Processing [cls, generation; PyTorch]
  • HGMR: Hierarchical Gaussian Mixtures for Adaptive 3D Registration [registration; unofficial code]
  • EC-Net: an Edge-aware Point set Consolidation Network [consolidation; Tensorflow]
  • Learning and Matching Multi-View Descriptors for Registration of Point Clouds [registration]
  • Local Spectral Graph Convolution for Point Set Feature Learning [cls, seg]
  • 3D Recurrent Neural Networks with Context Fusion for Point Cloud Semantic Segmentation [seg]
  • Fully-Convolutional Point Networks for Large-Scale Point Clouds [seg, captioning; Tensorflow]
  • PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptors [registration; PyTorch-unofficial]
  • Deep Continuous Fusion for Multi-Sensor 3D Object Detection [det]
  • 3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration [match, registration; Tensorflow]
  • Stereo Vision-based Semantic 3D Object and Ego-motion Tracking for Autonomous Driving [autonomous driving]
• Others
  • PointCNN: Convolution On X -Transformed Points [cls, seg; Tensorflow; NeurIPS]
  • Learning Representations and Generative Models for 3D Point Clouds [autoencoder; Tensorflow; ICML]
  • RGCNN: Regularized Graph CNN for Point Cloud Segmentation [seg, cls; Tensorflow; ACM MM]
  • PCN: Point Completion Network [completion; Tensorflow; 3DV]
  • Iterative Global Similarity Points : A robust coarse-to-fine integration solution for pairwise 3D point cloud registration [registration; 3DV]
  • Semantic Classification of 3D Point Clouds with Multiscale Spherical Neighborhoods [seg; 3DV]
  • Guaranteed Outlier Removal for Point Cloud Registration with Correspondences [registration; TPAMI]
  • Second: Sparsely embedded convolutional detection [det; Sensors]
  • Rt3d: Real-time 3-d vehicle detection in lidar point cloud for autonomous driving [det, autonomous driving; IEEE Robotics and Automation Letters]
  • HDNET: Exploiting HD Maps for 3D Object Detection [det, autonomous driving; CoRL]
  • Joint 3D Proposal Generation and Object Detection from View Aggregation [det, autonomous driving; IROS]
  • Flex-Convolution(Million-Scale Point-Cloud Learning Beyond Grid-Worlds) [cls, seg; Tensorflow; ACCV]
  • SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloud [seg; Tensorflow; ICRA]
  • Monte Carlo Convolution for Learning on Non-Uniformly Sampled Point Clouds [seg, cls, normal estimation; Tensorflow; TOG]
  • Learning Efficient Point Cloud Generation for Dense 3D Object Reconstruction [reconstruction; Tensorflow; AAAI]
• arXiv
  • Spherical Convolutional Neural Network for 3D Point Clouds [cls]
  • Point Convolutional Neural Networks by Extension Operators [cls, seg, normal estimation; Tensorflow]
  • PointSIFT: A SIFT-like Network Module for 3D Point Cloud Semantic Segmentation [seg; Tensorflow]
  • Point Cloud GAN [generation; PyTorch]
  • Roarnet: A robust 3d object detection based on region approximation refinement [det]
  • Classification of Point Cloud Scenes with Multiscale Voxel Deep Network [seg]

2017

• CVPR
  • Fine-To-Coarse Global Registration of RGB-D Scans [registration; Github]
  • Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesis [completion; Torch7]
  • SyncSpecCNN: Synchronized Spectral CNN for 3D Shape Segmentation [seg, keypoints; Github]
  • A Point Set Generation Network for 3D Object Reconstruction From a Single Image [reconstruction; Tensorflow]
  • Multi-View 3D Object Detection Network for Autonomous Driving [det, autonomous driving; Tensorflow]
  • Deep MANTA: A Coarse-To-Fine Many-Task Network for Joint 2D and 3D Vehicle Analysis From Monocular Image [autonomous driving]
  • PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation [cls, seg; Tensorflow]
  • 3D Bounding Box Estimation Using Deep Learning and Geometry [det]
  • OctNet: Learning Deep 3D Representations at High Resolutions [cls, seg, orientation estimation; PyTorch]
  • 3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions [match, registration; project]
  • 3D Point Cloud Registration for Localization using a Deep Neural Network Auto-Encoder [registration; github]
• ICCV
  • High-Resolution Shape Completion Using Deep Neural Networks for Global Structure and Local Geometry Inference [completion]
  • Escape from Cells: Deep Kd-Networks for the Recognition of 3D Point Cloud Models [cls, retrieval, seg; PyTorch-unofficial]
  • Learning Compact Geometric Features [registration; Github]
  • 2D-Driven 3D Object Detection in RGB-D Images [det]
• Others
  • Pointnet++: Deep hierarchical feature learning on point sets in a metric space [cls, seg; Tensorflow; NIPS]
  • Deep Sets [PyTorch; cls]
  • 3D Object Proposals Using Stereo Imagery for Accurate Object Class Detection [det, autonomous driving; TPAMI]
  • O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis [cls, retrieval, seg; Github; TOG]
  • Vote3deep: Fast object detection in 3d point clouds using efficient convolutional neural networks [det; ICRA]
  • 3d fully convolutional network for vehicle detection in point cloud [det; Tensorflow; IROS]
  • Shape Completion Enabled Robotic Grasping [completion; Keras; IROS]
  • SEGCloud: Semantic Segmentation of 3D Point Clouds [seg; 3DV]

Before 2016

• 2016
  • Fast Global Registration [registration; ECCV; Github]
  • Monocular 3D Object Detection for Autonomous Driving [CVPR]
  • Volumetric and Multi-View CNNs for Object Classification on 3D Data [CVPR]
  • Three-Dimensional Object Detection and Layout Prediction Using Clouds of Oriented Gradients [CVPR]
  • Deep Sliding Shapes for Amodal 3D Object Detection in RGB-D Images [CVPR]
  • Fpnn: Field probing neural networks for 3d data [NIPS]
  • Vehicle Detection from 3D Lidar Using Fully Convolutional Network [RSS]
• 2015
  • Robust Reconstruction of Indoor Scenes [reconstruction; CVPR]
  • Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration [registration; TPAMI; Github]
  • 3D ShapeNets: A Deep Representation for Volumetric Shapes [CVPR]
  • SUN RGB-D: A RGB-D Scene Understanding Benchmark Suite [CVPR]
  • Data-Driven 3D Voxel Patterns for Object Category Recognition [CVPR]
  • Multi-view convolutional neural networks for 3d shape recognition [ICCV]
  • 3d object proposals for accurate object class detection [NIPS]
  • Voting for Voting in Online Point Cloud Object [RSS]
  • Voxnet: A 3d convolutional neural network for real-time object recognition [IROS]
• 2014
  • Are Cars Just 3D Boxes? - Jointly Estimating the 3D Shape of Multiple Objects [CVPR]
  • Sliding shapes for 3d object detection in depth images [ECCV]
• 2013
  • Vision meets robotics: The kitti dataset [IJRR]
• 2012
  • Are we ready for autonomous driving? the kitti vision benchmark suite [CVPR]
• 2009
  • Fast point feature histograms (FPFH) for 3D registration [registration; ICRA]
  • Generalized-ICP [registration; RSS]
• 1992
  • A method for registration of 3-D shapes [registration; TPAMI]
• 1987
  • Least-squares fitting of two 3-D point sets [registration; TPAMI]

Resources

• https://github.com/Yochengliu/awesome-point-cloud-analysis
• https://github.com/yinyunie/3D-Shape-Analysis-Paper-List
• https://github.com/NUAAXQ/awesome-point-cloud-analysis-2020
• https://github.com/QingyongHu/SoTA-Point-Cloud
• https://github.com/timzhang642/3D-Machine-Learning
• https://github.com/XuyangBai/awesome-point-cloud-registration
• https://github.com/weiweisun2018/awesome-point-clouds-registration
• https://github.com/chaytonmin/Awesome-BEV-Perception-Multi-Cameras
• https://github.com/dragonlong/Trending-in-3D-Vision
• https://github.com/4DVLab/Vision-Centric-BEV-Perception
• https://github.com/autodriving-heart/Awesome-occupancy-perception

Tools

• Open3D: https://github.com/intel-isl/Open3D
• PCL: https://github.com/PointCloudLibrary/pcl
• PCL-Python: https://github.com/strawlab/python-pcl
• Torch-Points3D: https://github.com/nicolas-chaulet/torch-points3d
• mmdetection3d: https://github.com/open-mmlab/mmdetection3d
• OpenPCDet: https://github.com/open-mmlab/OpenPCDet
• PyTorch3D: https://github.com/facebookresearch/pytorch3d
• Minkowski Engine: https://github.com/NVIDIA/MinkowskiEngine
• pointcloudset: https://github.com/virtual-vehicle/pointcloudset

Software

• CloudCompare: https://github.com/cloudcompare/cloudcompare
• Semantic Segmentation Editor: https://github.com/Hitachi-Automotive-And-Industry-Lab/semantic-segmentation-editor
• MeshLab: https://github.com/cnr-isti-vclab/meshlab