mlsanityData to 7 Oct 2026

Beyond 3D Siamese Tracking: A Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds

arXiv 2203.01730 ↗First posted 3 Mar 2022cs.CVcited by 55

Not on the map. Why?

Abstract (arXiv)

3D single object tracking (3D SOT) in LiDAR point clouds plays a crucial role in autonomous driving. Current approaches all follow the Siamese paradigm based on appearance matching. However, LiDAR point clouds are usually textureless and incomplete, which hinders effective appearance matching. Besides, previous methods greatly overlook the critical motion clues among targets. In this work, beyond 3D Siamese tracking, we introduce a motion-centric paradigm to handle 3D SOT from a new perspective. Following this paradigm, we propose a matching-free two-stage tracker M^2-Track. At the 1^st-stage, M^2-Track localizes the target within successive frames via motion transformation. Then it refines the target box through motion-assisted shape completion at the 2^nd-stage. Extensive experiments confirm that M^2-Track significantly outperforms previous state-of-the-arts on three large-scale datasets while running at 57FPS (~8%, ~17%, and ~22%) precision gains on KITTI, NuScenes, and Waymo Open Dataset respectively). Further analysis verifies each component's effectiveness and shows the motion-centric paradigm's promising potential when combined with appearance matching.

Key facts. Cited by 55 papers in the map (data through 7 Oct 2026). Builds on 38 of its 46 references that are in the map. First posted on arXiv 3 Mar 2022.

Builds on

38 of 46 references.

References · 38

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  4. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  5. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  6. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,436
  7. 2014Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modelingcited by 2,341
  8. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  9. 2020Center-based 3D Object Detection and Trackingcited by 1,143
  10. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 939
  11. 2019Deep Hough Voting for 3D Object Detection in Point Cloudscited by 648
  12. 2018SiamRPN++: Evolution of Siamese Visual Tracking with Very Deep Networkscited by 445
26 more
  1. 2019From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Networkcited by 366
  2. 2019Learning Discriminative Model Prediction for Trackingcited by 344
  3. 2019Fooling automated surveillance cameras: adversarial patches to attack person detectioncited by 268
  4. 2020PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Samplingcited by 238
  5. 2020Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completioncited by 222
  6. 2021Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Trackingcited by 220
  7. 2018Distractor-aware Siamese Networks for Visual Object Trackingcited by 187
  8. 2019SiamFC++: Towards Robust and Accurate Visual Tracking with Target Estimation Guidelinescited by 177
  9. 2021Offboard 3D Object Detection from Point Cloud Sequencescited by 144
  10. 2020PnPNet: End-to-End Perception and Prediction with Tracking in the Loopcited by 143
  11. 2020Know Your Surroundings: Exploiting Scene Information for Object Trackingcited by 119
  12. 2020P2B: Point-to-Box Network for 3D Object Tracking in Point Cloudscited by 84
  13. 2019Leveraging Shape Completion for 3D Siamese Trackingcited by 74
  14. 2019What You See is What You Get: Exploiting Visibility for 3D Object Detectioncited by 69
  15. 2021Box-Aware Feature Enhancement for Single Object Tracking on Point Cloudscited by 66
  16. 2020LiDAR-based Online 3D Video Object Detection with Graph-based Message Passing and Spatiotemporal Transformer Attentioncited by 65
  17. 2020Probabilistic 3D Multi-Object Tracking for Autonomous Drivingcited by 64
  18. 2021PTT: Point-Track-Transformer Module for 3D Single Object Tracking in Point Cloudscited by 59
  19. 2021Exploring Simple 3D Multi-Object Tracking for Autonomous Drivingcited by 56
  20. 20213D Siamese Voxel-to-BEV Tracker for Sparse Point Cloudscited by 54
  21. 2020CaSPR: Learning Canonical Spatiotemporal Point Cloud Representationscited by 51
  22. 20213D Object Tracking with Transformercited by 49
  23. 2020An LSTM Approach to Temporal 3D Object Detection in LiDAR Point Cloudscited by 45
  24. 20213D-SiamRPN: An End-to-End Learning Method for Real-Time 3D Single Object Tracking Using Raw Point Cloudcited by 45
  25. 20214D-Net for Learned Multi-Modal Alignmentcited by 44
  26. 2019Efficient Bird Eye View Proposals for 3D Siamese Trackingcited by 18

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 7; 2023: 20; 2024: 19; 2025: 4; 2026: 5.

55 papers.

Sort:

  1. 20222DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Cloudscited by 157
  2. 2022CXTrack: Improving 3D Point Cloud Tracking with Contextual Informationcited by 28
  3. 2023MBPTrack: Improving 3D Point Cloud Tracking with Memory Networks and Box Priorscited by 24
  4. 2023DropMAE: Learning Representations via Masked Autoencoders with Spatial-Attention Dropout for Temporal Matching Taskscited by 22
  5. 2022Exploring Point-BEV Fusion for 3D Point Cloud Object Tracking with Transformercited by 18
  6. 2023Human-centric Scene Understanding for 3D Large-scale Scenarioscited by 17
  7. 2023OSP2B: One-Stage Point-to-Box Network for 3D Siamese Trackingcited by 16
  8. 2023Synchronize Feature Extracting and Matching: A Single Branch Framework for 3D Object Trackingcited by 16
  9. 2022Exploiting More Information in Sparse Point Cloud for 3D Single Object Trackingcited by 14
  10. 2023An Effective Motion-Centric Paradigm for 3D Single Object Tracking in Point Cloudscited by 14
  11. 2024Towards Category Unification of 3D Single Object Tracking on Point Cloudscited by 13
  12. 2023Correlation Pyramid Network for 3D Single Object Trackingcited by 10
38 more
  1. 2023STTracker: Spatio-Temporal Tracker for 3D Single Object Trackingcited by 10
  2. 2023M3SOT: Multi-frame, Multi-field, Multi-space 3D Single Object Trackingcited by 10
  3. 2024P2P: Part-to-Part Motion Cues Guide a Strong Tracking Framework for LiDAR Point Cloudscited by 10
  4. 2024MVCTrack: Boosting 3D Point Cloud Tracking via Multimodal-Guided Virtual Cuescited by 10
  5. 2024RobMOT: Robust 3D Multi-Object Tracking by Observational Noise and State Estimation Drift Mitigation on LiDAR PointCloudcited by 9
  6. 2025CompTrack: Information Bottleneck-Guided Low-Rank Dynamic Token Compression for Point Cloud Trackingcited by 9
  7. 2023Modeling Continuous Motion for 3D Point Cloud Object Trackingcited by 8
  8. 2023An Efficient Semi-Automated Scheme for Infrastructure LiDAR Annotationcited by 7
  9. 2023Robust 3D Tracking with Quality-Aware Shape Completioncited by 7
  10. 2026FocusTrack: One-Stage Focus-and-Suppress Framework for 3D Point Cloud Object Trackingcited by 7
  11. 2022Implicit and Efficient Point Cloud Completion for 3D Single Object Trackingcited by 6
  12. 2024SeqTrack3D: Exploring Sequence Information for Robust 3D Point Cloud Trackingcited by 6
  13. 2022OST: Efficient One-stream Network for 3D Single Object Tracking in Point Cloudscited by 5
  14. 2024VoxelTrack: Exploring Voxel Representation for 3D Point Cloud Object Trackingcited by 5
  15. 20243D Single-object Tracking in Point Clouds with High Temporal Variationcited by 4
  16. 2025TrackAny3D: Transferring Pretrained 3D Models for Category-unified 3D Point Cloud Trackingcited by 4
  17. 2025Beyond Frame-wise Tracking: A Trajectory-based Paradigm for Efficient Point Cloud Trackingcited by 4
  18. 2022Emerging Threats in Deep Learning-Based Autonomous Driving: A Comprehensive Surveycited by 3
  19. 2023MixCycle: Mixup Assisted Semi-Supervised 3D Single Object Tracking with Cycle Consistencycited by 3
  20. 2023Motion-to-Matching: A Mixed Paradigm for 3D Single Object Trackingcited by 3
  21. 2024BirdNeRF: Fast Neural Reconstruction of Large-Scale Scenes From Aerial Imagerycited by 3
  22. 2024FlowTrack: Point-level Flow Network for 3D Single Object Trackingcited by 3
  23. 2023BEVTrack: A Simple and Strong Baseline for 3D Single Object Tracking in Bird's-Eye Viewcited by 2
  24. 2024PillarTrack:Boosting Pillar Representation for Transformer-based 3D Single Object Tracking on Point Cloudscited by 2
  25. 2024GSOT3D: Towards Generic 3D Single Object Tracking in the Wildcited by 2
  26. 2026TFTrack: A Template-Free Framework for Efficient 3D Point Cloud Trackingcited by 2
  27. 2023Object Preserving Siamese Network for Single Object Tracking on Point Cloudscited by 1
  28. 2023MMF-Track: Multi-modal Multi-level Fusion for 3D Single Object Trackingcited by 1
  29. 2023Hierarchical Adaptive Voxel-guided Sampling for Real-time Applications in Large-scale Point Cloudscited by 1
  30. 2023Multi-Correlation Siamese Transformer Network with Dense Connection for 3D Single Object Trackingcited by 1
  31. 2024Small Object Tracking in LiDAR Point Cloud: Learning the Target-awareness Prototype and Fine-grained Search Regioncited by 1
  32. 2024AYDIV: Adaptable Yielding 3D Object Detection via Integrated Contextual Vision Transformercited by 1
  33. 2024D-Aug: Enhancing Data Augmentation for Dynamic LiDAR Scenescited by 1
  34. 2024SiamMo: Siamese Motion-Centric 3D Object Trackingcited by 1
  35. 2023Adaptive Local Adversarial Attacks on 3D Point Clouds for Augmented Realitycited by 0
  36. 2024Spatio-Temporal Bi-directional Cross-frame Memory for Distractor Filtering Point Cloud Single Object Trackingcited by 0
  37. 2024EasyTrack: Efficient and Compact One-stream 3D Point Clouds Trackercited by 0
  38. 2024Quantum Inverse Contextual Vision Transformers (Q-ICVT): A New Frontier in 3D Object Detection for AVscited by 0

50 most cited of 55 All 55 (JSON)