mlsanityData to 14 Oct 2026

PolarMOT: How Far Can Geometric Relations Take Us in 3D Multi-Object Tracking?

arXiv 2208.01957 ↗First posted 3 Aug 2022cs.CV cs.LGcited by 29

Not on the map. Why?

Abstract (arXiv)

Most (3D) multi-object tracking methods rely on appearance-based cues for data association. By contrast, we investigate how far we can get by only encoding geometric relationships between objects in 3D space as cues for data-driven data association. We encode 3D detections as nodes in a graph, where spatial and temporal pairwise relations among objects are encoded via localized polar coordinates on graph edges. This representation makes our geometric relations invariant to global transformations and smooth trajectory changes, especially under non-holonomic motion. This allows our graph neural network to learn to effectively encode temporal and spatial interactions and fully leverage contextual and motion cues to obtain final scene interpretation by posing data association as edge classification. We establish a new state-of-the-art on nuScenes dataset and, more importantly, show that our method, PolarMOT, generalizes remarkably well across different locations (Boston, Singapore, Karlsruhe) and datasets (nuScenes and KITTI).

Key facts. Cited by 29 papers in the map (data through 14 Oct 2026). Builds on 36 of its 67 references that are in the map. First posted on arXiv 3 Aug 2022.

Builds on

36 of 67 references.

References · 36

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  3. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  4. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  5. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,401
  6. 2017Neural Message Passing for Quantum Chemistrycited by 3,487
  7. 2018PointPillars: Fast Encoders for Object Detection from Point Cloudscited by 1,715
  8. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  9. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,205
  10. 2020Center-based 3D Object Detection and Trackingcited by 1,144
  11. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 941
  12. 2016Simple Online and Realtime Trackingcited by 868
24 more
  1. 2019On the Variance of the Adaptive Learning Rate and Beyondcited by 772
  2. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
  3. 2020HOTA: A Higher Order Metric for Evaluating Multi-Object Trackingcited by 552
  4. 2020Tracking Objects as Pointscited by 483
  5. 2019Tracking without bells and whistlescited by 382
  6. 2020Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloudcited by 335
  7. 2019MOTS: Multi-Object Tracking and Segmentationcited by 223
  8. 2019Learning a Neural Solver for Multiple Object Trackingcited by 198
  9. 2020Video Panoptic Segmentationcited by 142
  10. 2021Learning to Track with Object Permanencecited by 119
  11. 2020GNN3DMOT: Graph Neural Network for 3D Multi-Object Tracking with Multi-Feature Learningcited by 113
  12. 2021EagerMOT: 3D Multi-Object Tracking via Sensor Fusioncited by 83
  13. 2016Learning by tracking: Siamese CNN for robust target associationcited by 79
  14. 2019How To Train Your Deep Multi-Object Trackercited by 71
  15. 20214D Panoptic LiDAR Segmentationcited by 65
  16. 2020Probabilistic 3D Multi-Object Tracking for Autonomous Drivingcited by 64
  17. 2019Track to Reconstruct and Reconstruct to Trackcited by 49
  18. 2015Near-Online Multi-target Tracking with Aggregated Local Flow Descriptorcited by 46
  19. 2017Deep Network Flow for Multi-Object Trackingcited by 45
  20. 2018Beyond Pixels: Leveraging Geometry and Shape Cues for Online Multi-Object Trackingcited by 40
  21. 2021Learnable Online Graph Representations for 3D Multi-Object Trackingcited by 38
  22. 2018Combined Image- and World-Space Tracking in Traffic Scenescited by 29
  23. 2018End-to-end Learning of Multi-sensor 3D Tracking by Detectioncited by 25
  24. 2017Track, then Decide: Category-Agnostic Vision-based Multi-Object Trackingcited by 19

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 11; 2024: 9; 2025: 7; 2026: 2.

29 papers.

Sort:

By area:

  1. 2023ReST: A Reconfigurable Spatial-Temporal Graph Model for Multi-Camera Multi-Object Trackingcited by 24
  2. 20233DMOTFormer: Graph Transformer for Online 3D Multi-Object Trackingcited by 21
  3. 2023RaTrack: Moving Object Detection and Tracking with 4D Radar Point Cloudcited by 21
  4. 2023ByteTrackV2: 2D and 3D Multi-Object Tracking by Associating Every Detection Boxcited by 20
  5. 2023TrajectoryFormer: 3D Object Tracking Transformer with Predictive Trajectory Hypothesescited by 14
  6. 2023LEGO: Learning and Graph-Optimized Modular Tracker for Online Multi-Object Tracking with Point Cloudscited by 9
  7. 2024PNAS-MOT: Multi-Modal Object Tracking with Pareto Neural Architecture Searchcited by 9
  8. 2024RobMOT: Robust 3D Multi-Object Tracking by Observational Noise and State Estimation Drift Mitigation on LiDAR PointCloudcited by 9
  9. 2023You Only Need Two Detectors to Achieve Multi-Modal 3D Multi-Object Trackingcited by 8
  10. 2025Zero-Shot 4D Lidar Panoptic Segmentationcited by 6
  11. 20233D Multiple Object Tracking on Autonomous Driving: A Literature Reviewcited by 4
  12. 20233D Multi-Object Tracking Based on Uncertainty-Guided Data Associationcited by 3
17 more
  1. 2025HybridTrack: A Hybrid Approach for Robust Multi-Object Trackingcited by 3
  2. 2023DecoderTracker: Decoder-Only Method for Multiple-Object Trackingcited by 2
  3. 2024InScope: A New Real-world 3D Infrastructure-side Collaborative Perception Dataset for Open Traffic Scenarioscited by 2
  4. 2024CrossTracker: Robust Multi-modal 3D Multi-Object Tracking via Cross Correctioncited by 2
  5. 2025One Graph to Track Them All: Dynamic GNNs for Single- and Multi-View Trackingcited by 2
  6. 2023BOTT: Box Only Transformer Tracker for 3D Object Trackingcited by 1
  7. 2024DeepKalPose: An Enhanced Deep-Learning Kalman Filter for Temporally Consistent Monocular Vehicle Pose Estimationcited by 1
  8. 2024STT: Stateful Tracking with Transformers for Autonomous Drivingcited by 1
  9. 2024MonoDETRNext: Next-Generation Accurate and Efficient Monocular 3D Object Detectorcited by 1
  10. 2024Spb3DTracker: A Robust LiDAR-Based Person Tracker for Noisy Environmentcited by 1
  11. 2025OptiPMB: Enhancing 3D Multi-Object Tracking with Optimized Poisson Multi-Bernoulli Filteringcited by 1
  12. 2025Towards Accurate State Estimation: Motion Dynamics Kalman Filter for 3D Multi-Object Trackingcited by 1
  13. 2025ATLASFusion: Aggregation Tracking with Location-Aware Sparse Fusion for Robust Spatio-Temporal Multi-View Pedestrian Trackingcited by 1
  14. 2026Polycepta: Object-Centric Appearance Estimation for Multi-Object Trackingcited by 1
  15. 2024RegTrack: Simplicity Beneath Complexity in Robust Multi-Modal 3D Multi-Object Trackingcited by 0
  16. 2025S3MOT: Monocular 3D Object Tracking with Selective State Space Modelcited by 0
  17. 2026Beyond Leaderboard Scores: A Deployment-Focused Protocol for Interpretable Tracking Evaluation in Pedestrian-Centric Environmentscited by 0