mlsanityData to 7 Oct 2026

SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generation

arXiv 2108.06709 ↗First posted 15 Aug 2021cs.CVcited by 93

Not on the map. Why?

Abstract (arXiv)

In autonomous driving, a LiDAR-based object detector should perform reliably at different geographic locations and under various weather conditions. While recent 3D detection research focuses on improving performance within a single domain, our study reveals that the performance of modern detectors can drop drastically cross-domain. In this paper, we investigate unsupervised domain adaptation (UDA) for LiDAR-based 3D object detection. On the Waymo Domain Adaptation dataset, we identify the deteriorating point cloud quality as the root cause of the performance drop. To address this issue, we present Semantic Point Generation (SPG), a general approach to enhance the reliability of LiDAR detectors against domain shifts. Specifically, SPG generates semantic points at the predicted foreground regions and faithfully recovers missing parts of the foreground objects, which are caused by phenomena such as occlusions, low reflectance or weather interference. By merging the semantic points with the original points, we obtain an augmented point cloud, which can be directly consumed by modern LiDAR-based detectors. To validate the wide applicability of SPG, we experiment with two representative detectors, PointPillars and PV-RCNN. On the UDA task, SPG significantly improves both detectors across all object categories of interest and at all difficulty levels. SPG can also benefit object detection in the original domain. On the Waymo Open Dataset and KITTI, SPG improves 3D detection results of these two methods across all categories. Combined with PV-RCNN, SPG achieves state-of-the-art 3D detection results on KITTI.

Key facts. Cited by 93 papers in the map (data through 7 Oct 2026). Builds on 62 of its 79 references that are in the map. First posted on arXiv 15 Aug 2021.

Builds on

62 of 79 references.

References · 62

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  2. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,436
  3. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  4. 2018PointPillars: Fast Encoders for Object Detection from Point Cloudscited by 1,707
  5. 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,323
  6. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  7. 2017Adversarial Discriminative Domain Adaptationcited by 1,309
  8. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 939
  9. 2019PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detectioncited by 890
  10. 2016Multi-View 3D Object Detection Network for Autonomous Drivingcited by 721
  11. 2017Maximum Classifier Discrepancy for Unsupervised Domain Adaptationcited by 682
  12. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
38 more
  1. 2020Contrastive Learning for Unpaired Image-to-Image Translationcited by 637
  2. 2018Domain Adaptive Faster R-CNN for Object Detection in the Wildcited by 442
  3. 20203DSSD: Point-based 3D Single Stage Object Detectorcited by 440
  4. 2020Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detectioncited by 409
  5. 2018PCN: Point Completion Networkcited by 382
  6. 2019From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Networkcited by 366
  7. 2019PIXOR: Real-time 3D Object Detection from Point Cloudscited by 358
  8. 2016Unsupervised Domain Adaptation with Residual Transfer Networkscited by 340
  9. 2020Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloudcited by 335
  10. 2017Joint 3D Proposal Generation and Object Detection from View Aggregationcited by 325
  11. 2019STD: Sparse-to-Dense 3D Object Detector for Point Cloudcited by 282
  12. 2020Deep Continuous Fusion for Multi-Sensor 3D Object Detectioncited by 263
  13. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  14. 2018PU-Net: Point Cloud Upsampling Networkcited by 208
  15. 2020Multi-Task Multi-Sensor Fusion for 3D Object Detectioncited by 203
  16. 20203D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View Spatial Feature Fusion for 3D Object Detectioncited by 176
  17. 2020GRNet: Gridding Residual Network for Dense Point Cloud Completioncited by 174
  18. 2019End-to-End Multi-View Fusion for 3D Object Detection in LiDAR Point Cloudscited by 171
  19. 2020EPNet: Enhancing Point Features with Image Semantics for 3D Object Detectioncited by 164
  20. 2019Stereo R-CNN based 3D Object Detection for Autonomous Drivingcited by 163
  21. 2019PU-GAN: a Point Cloud Upsampling Adversarial Networkcited by 162
  22. 2020Train in Germany, Test in The USA: Making 3D Object Detectors Generalizecited by 150
  23. 2020CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detectioncited by 147
  24. 2019Progressive Domain Adaptation for Object Detectioncited by 134
  25. 2019LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Drivingcited by 133
  26. 2019IoU Loss for 2D/3D Object Detectioncited by 128
  27. 2019Self-Ensembling with GAN-based Data Augmentation for Domain Adaptation in Semantic Segmentationcited by 126
  28. 2019Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detectioncited by 124
  29. 2019Fast Point R-CNNcited by 119
  30. 2019PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud Representationcited by 114
  31. 2019Multi-adversarial Faster-RCNN for Unrestricted Object Detectioncited by 111
  32. 2020CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloudcited by 107
  33. 2019GS3D: An Efficient 3D Object Detection Framework for Autonomous Drivingcited by 106
  34. 2019Grid-GCN for Fast and Scalable Point Cloud Learningcited by 103
  35. 2018Patch-based Progressive 3D Point Set Upsamplingcited by 101
  36. 2020DSGN: Deep Stereo Geometry Network for 3D Object Detectioncited by 100
  37. 2019Object as Hotspots: An Anchor-Free 3D Object Detection Approach via Firing of Hotspotscited by 84
  38. 2020Range Conditioned Dilated Convolutions for Scale Invariant 3D Object Detectioncited by 78

50 most cited of 62 All 62 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 5; 2022: 27; 2023: 28; 2024: 16; 2025: 10; 2026: 7.

93 papers.

Sort:

By area:

  1. 2022Focal Sparse Convolutional Networks for 3D Object Detectioncited by 142
  2. 20223D Object Detection for Autonomous Driving: A Comprehensive Surveycited by 132
  3. 2021The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by Normalizationcited by 131
  4. 2022Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completioncited by 82
  5. 2023UltraLiDAR: Learning Compact Representations for LiDAR Completion and Generationcited by 76
  6. 2022Point Density-Aware Voxels for LiDAR 3D Object Detectioncited by 70
  7. 2021Behind the Curtain: Learning Occluded Shapes for 3D Object Detectioncited by 58
  8. 2022LiDAR Distillation: Bridging the Beam-Induced Domain Gap for 3D Object Detectioncited by 54
  9. 20233D Semantic Segmentation in the Wild: Learning Generalized Models for Adverse-Condition Point Cloudscited by 46
  10. 2022Graph R-CNN: Towards Accurate 3D Object Detection with Semantic-Decorated Local Graphcited by 42
  11. 2022Domain Adaptive Object Detection for Autonomous Driving under Foggy Weathercited by 41
  12. 2023Uni3D: A Unified Baseline for Multi-dataset 3D Object Detectioncited by 39
38 more
  1. 2022SSDA3D: Semi-supervised Domain Adaptation for 3D Object Detection from Point Cloudcited by 37
  2. 2024L4DR: LiDAR-4DRadar Fusion for Weather-Robust 3D Object Detectioncited by 32
  3. 2023Density-Insensitive Unsupervised Domain Adaption on 3D Object Detectioncited by 31
  4. 2022Deformation and Correspondence Aware Unsupervised Synthetic-to-Real Scene Flow Estimation for Point Cloudscited by 28
  5. 2023Revisiting Domain-Adaptive 3D Object Detection by Reliable, Diverse and Class-balanced Pseudo-Labelingcited by 28
  6. 2021DSOR: A Scalable Statistical Filter for Removing Falling Snow from LiDAR Point Clouds in Severe Winter Weathercited by 26
  7. 2022CL3D: Unsupervised Domain Adaptation for Cross-LiDAR 3D Detectioncited by 26
  8. 2022GLENet: Boosting 3D Object Detectors with Generative Label Uncertainty Estimationcited by 24
  9. 2023Bi3D: Bi-domain Active Learning for Cross-domain 3D Object Detectioncited by 21
  10. 2023SALUDA: Surface-based Automotive Lidar Unsupervised Domain Adaptationcited by 21
  11. 2023PVT-SSD: Single-Stage 3D Object Detector with Point-Voxel Transformercited by 21
  12. 2023A Survey of Label-Efficient Deep Learning for 3D Point Cloudscited by 20
  13. 2023S2R-ViT for Multi-Agent Cooperative Perception: Bridging the Gap from Simulation to Realitycited by 18
  14. 2022Dense Voxel Fusion for 3D Object Detectioncited by 17
  15. 2022DeepFusion: A Robust and Modular 3D Object Detector for Lidars, Cameras and Radarscited by 16
  16. 2021See Eye to Eye: A Lidar-Agnostic 3D Detection Framework for Unsupervised Multi-Target Domain Adaptationcited by 15
  17. 2022Sparse2Dense: Learning to Densify 3D Features for 3D Object Detectioncited by 14
  18. 2023A Survey on Approximate Edge AI for Energy Efficient Autonomous Driving Servicescited by 14
  19. 2023MDT3D: Multi-Dataset Training for LiDAR 3D Object Detection Generalizationcited by 14
  20. 2022Unsupervised RGB-to-Thermal Domain Adaptation via Multi-Domain Attention Networkcited by 13
  21. 2023Unsupervised Adaptation from Repeated Traversals for Autonomous Drivingcited by 13
  22. 2023ReSimAD: Zero-Shot 3D Domain Transfer for Autonomous Driving with Source Reconstruction and Target Simulationcited by 13
  23. 2024DG-PIC: Domain Generalized Point-In-Context Learning for Point Cloud Understandingcited by 13
  24. 2024Unified Domain Generalization and Adaptation for Multi-View 3D Object Detectioncited by 13
  25. 2023PG-RCNN: Semantic Surface Point Generation for 3D Object Detectioncited by 12
  26. 2024UADA3D: Unsupervised Adversarial Domain Adaptation for 3D Object Detection with Sparse LiDAR and Large Domain Gapscited by 12
  27. 2024Sunshine to Rainstorm: Cross-Weather Knowledge Distillation for Robust 3D Object Detectioncited by 11
  28. 2022Multimodal Transformer for Automatic 3D Annotation and Object Detectioncited by 10
  29. 2022Viewer-Centred Surface Completion for Unsupervised Domain Adaptation in 3D Object Detectioncited by 10
  30. 2023Deep Transfer Learning for Intelligent Vehicle Perception: a Surveycited by 10
  31. 2024An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domainscited by 10
  32. 2022CARLANE: A Lane Detection Benchmark for Unsupervised Domain Adaptation from Simulation to multiple Real-World Domainscited by 9
  33. 2023DualCross: Cross-Modality Cross-Domain Adaptation for Monocular BEV Perceptioncited by 8
  34. 2024SOAP: Cross-sensor Domain Adaptation for 3D Object Detection Using Stationary Object Aggregation Pseudo-labellingcited by 8
  35. 2024CMDA: Cross-Modal and Domain Adversarial Adaptation for LiDAR-Based 3D Object Detectioncited by 8
  36. 2023Domain Adaptation based Object Detection for Autonomous Driving in Foggy and Rainy Weathercited by 7
  37. 2025R2LDM: An Efficient 4D Radar Super-Resolution Framework Leveraging Diffusion Modelcited by 7
  38. 2022Boosting Single-Frame 3D Object Detection by Simulating Multi-Frame Point Cloudscited by 6

50 most cited of 93 All 93 (JSON)