mlsanityData to 14 Oct 2026

ByteTrackV2: 2D and 3D Multi-Object Tracking by Associating Every Detection Box

arXiv 2303.15334 ↗First posted 27 Mar 2023cs.CVcited by 20

Not on the map. Why?

Abstract (arXiv)

Multi-object tracking (MOT) aims at estimating bounding boxes and identities of objects across video frames. Detection boxes serve as the basis of both 2D and 3D MOT. The inevitable changing of detection scores leads to object missing after tracking. We propose a hierarchical data association strategy to mine the true objects in low-score detection boxes, which alleviates the problems of object missing and fragmented trajectories. The simple and generic data association strategy shows effectiveness under both 2D and 3D settings. In 3D scenarios, it is much easier for the tracker to predict object velocities in the world coordinate. We propose a complementary motion prediction strategy that incorporates the detected velocities with a Kalman filter to address the problem of abrupt motion and short-term disappearing. ByteTrackV2 leads the nuScenes 3D MOT leaderboard in both camera (56.4% AMOTA) and LiDAR (70.1% AMOTA) modalities. Furthermore, it is nonparametric and can be integrated with various detectors, making it appealing in real applications. The source code is released at https://github.com/ifzhang/ByteTrack-V2.

Key facts. Cited by 20 papers in the map (data through 14 Oct 2026). Builds on 107 of its 132 references that are in the map. First posted on arXiv 27 Mar 2023.

Builds on

107 of 132 references.

References · 107

  1. 2017Attention Is All You Needcited by 60,361
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  4. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  5. 2020End-to-End Object Detection with Transformerscited by 7,414
  6. 2017Mask R-CNNcited by 7,277
  7. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  8. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,523
  9. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,422
  10. 2018YOLOv3: An Incremental Improvementcited by 2,659
  11. 2018PointPillars: Fast Encoders for Object Detection from Point Cloudscited by 1,715
  12. 2020YOLOv4: Optimal Speed and Accuracy of Object Detectioncited by 1,630
38 more
  1. 2019Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regressioncited by 1,368
  2. 2017VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detectioncited by 1,323
  3. 2022DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detectioncited by 1,319
  4. 2018BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learningcited by 1,288
  5. 2017Cascade R-CNN: Delving into High Quality Object Detectioncited by 1,276
  6. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformerscited by 1,218
  7. 2019Objects as Pointscited by 1,175
  8. 2017Mixed Precision Trainingcited by 1,156
  9. 2020Center-based 3D Object Detection and Trackingcited by 1,144
  10. 2021YOLOX: Exceeding YOLO Series in 2021cited by 1,071
  11. 2020Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by Implicitly Unprojecting to 3Dcited by 1,044
  12. 2022YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectorscited by 982
  13. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 941
  14. 2017Simple Online and Realtime Tracking with a Deep Association Metriccited by 910
  15. 2016Simple Online and Realtime Trackingcited by 868
  16. 2021ByteTrack: Multi-Object Tracking by Associating Every Detection Boxcited by 823
  17. 2016Performance Measures and a Data Set for Multi-Target, Multi-Camera Trackingcited by 692
  18. 2021DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queriescited by 658
  19. 2022BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detectioncited by 572
  20. 2020HOTA: A Higher Order Metric for Evaluating Multi-Object Trackingcited by 552
  21. 2022PETR: Position Embedding Transformation for Multi-View 3D Object Detectioncited by 552
  22. 2016MOT16: A Benchmark for Multi-Object Trackingcited by 532
  23. 2021TrackFormer: Multi-Object Tracking with Transformerscited by 518
  24. 2020Sparse R-CNN: End-to-End Object Detection with Learnable Proposalscited by 495
  25. 2020FairMOT: On the Fairness of Detection and Re-Identification in Multiple Object Trackingcited by 490
  26. 2020Tracking Objects as Pointscited by 483
  27. 2022TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with Transformerscited by 466
  28. 2018Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Drivingcited by 457
  29. 2021FCOS3D: Fully Convolutional One-Stage Monocular 3D Object Detectioncited by 447
  30. 2021Conditional DETR for Fast Training Convergencecited by 424
  31. 2019Tracking without bells and whistlescited by 382
  32. 2021MOTR: End-to-End Multiple-Object Tracking with Transformercited by 381
  33. 2020TransTrack: Multiple Object Tracking with Transformercited by 371
  34. 2020MOT20: A benchmark for multi object tracking in crowded scenescited by 357
  35. 2022PETRv2: A Unified Framework for 3D Perception from Multi-Camera Imagescited by 352
  36. 2021Categorical Depth Distribution Network for Monocular 3D Object Detectioncited by 327
  37. 2019Towards Real-Time Multi-Object Trackingcited by 311
  38. 2018CrowdHuman: A Benchmark for Detecting Human in a Crowdcited by 284

50 most cited of 107 All 107 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 3; 2024: 10; 2025: 5; 2026: 2.

20 papers.

Sort:

By area:

  1. 2023DrivingDiffusion: Layout-Guided multi-view driving scene video generation with latent diffusion modelcited by 75
  2. 2024Unleashing HyDRa: Hybrid Fusion, Depth Consistency and Radar for Unified 3D Perceptioncited by 35
  3. 2024ADA-Track++: End-to-End Multi-Camera 3D Multi-Object Tracking with Alternating Detection and Associationcited by 33
  4. 2024MCTrack: A Unified 3D Multi-Object Tracking Framework for Autonomous Drivingcited by 17
  5. 2024Fast-Poly: A Fast Polyhedral Framework For 3D Multi-Object Trackingcited by 11
  6. 2024Seeing Beyond Views: Multi-View Driving Scene Video Generation with Holistic Attentioncited by 11
  7. 2023LEGO: Learning and Graph-Optimized Modular Tracker for Online Multi-Object Tracking with Point Cloudscited by 9
  8. 2024Radar and Camera Fusion for Object Detection and Tracking: A Comprehensive Surveycited by 9
  9. 2025Bridging Perspectives: Foundation Model Guided BEV Maps for 3D Object Detection and Trackingcited by 9
  10. 2024Beyond Kalman Filters: Deep Learning-Based Filters for Improved Object Trackingcited by 8
  11. 2024SpaRC: Sparse Radar-Camera Fusion for 3D Object Detectioncited by 8
  12. 20233D Multiple Object Tracking on Autonomous Driving: A Literature Reviewcited by 4
8 more
  1. 2024Engineering an Efficient Object Tracker for Non-Linear Motioncited by 4
  2. 2025Deep Learning-Based Multi-Object Tracking: A Comprehensive Survey from Foundations to State-of-the-Artcited by 4
  3. 2024RockTrack: A 3D Robust Multi-Camera-Ken Multi-Object Tracking Frameworkcited by 2
  4. 2025OptiPMB: Enhancing 3D Multi-Object Tracking with Optimized Poisson Multi-Bernoulli Filteringcited by 1
  5. 2025Delving into Dynamic Scene Cue-Consistency for Robust 3D Multi-Object Trackingcited by 0
  6. 2025Radar-Camera Fused Multi-Object Tracking: Online Calibration and Common Featurecited by 0
  7. 2026MatchFusion: Explicit-Implicit Instance Matching for Spatio-Temporal Multimodal Autonomous Drivingcited by 0
  8. 2026ByteTraX: Enhancing the ByteTrack Architecture with Optimised Thresholdingcited by 0