Pith. sign in

REVIEW 4 major objections 5 minor 42 references

MVCTrack: Boosting 3D Point Cloud Tracking via Multimodal-Guided Virtual Cues

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read Image-derived virtual points lift LiDAR-only 3D tracking accuracy on sparse scenes.

desk verdict A plausible, incremental transfer of virtual-point generation to 3D single-object tracking; the main uncertainty is the undisclosed mask source, which a careful revision should resolve. read the letter →

arxiv 2412.02734 v5 pith:AS3UZXIO submitted 2024-12-03 cs.CV cs.RO

classification cs.CVcs.RO
keywords 3DsingleobjecttrackingLiDARpointcloudmultimodalfusionvirtualcuesRGB-depthprojectionsparsenuScenesreal-time
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that 3D single-object tracking in sparse LiDAR point clouds can be improved by generating dense virtual points from 2D image segmentations and feeding them together with raw LiDAR points into an otherwise standard tracker. On the nuScenes benchmark, the proposed MVCTrack reaches mean success/precision of 61.20/73.22, surpassing its LiDAR-only baseline by 1.98/2.03 points, and the same virtual-cue injection also improves a second motion-based tracker. If the claim holds, cheap RGB sensors can act as a plug-in densifier for LiDAR-only 3D trackers, with the largest benefits for small and distant objects.

What carries the argument

The load-bearing mechanism is virtual-cue projection: a lightweight 2D segmentor produces instance masks; a fixed number $\tau$ of pixels are sampled uniformly inside each mask; each sampled pixel receives the depth of the nearest LiDAR point projected into the image (Eq. 5); the pixel with that depth is unprojected into a 3D virtual cue; and the virtual cues are concatenated with raw LiDAR points to form the augmented cloud $P_{\text{aug}} = P \cup V$. The ablation identifies the fixed-number uniform sampling strategy as the best choice, and the augmented cloud is fed directly into a standard 3D sparse-convolution tracker, so the scheme requires no change to the tracker's architecture.

What would settle it

Re-run the pipeline on nuScenes feeding masks from a real-time learned 2D segmentor instead of the masks used in the paper; if the margin over the 59.22/71.19 LiDAR-only baseline shrinks to near zero, the claim that virtual cues robustly drive the improvement is falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that a Multimodal-guided Virtual Cues Projection (MVCP) scheme, which samples pixels inside 2D instance masks, assigns each pixel the depth of its nearest projected LiDAR point, and unprojects it into a 3D virtual cue, produces an augmented point cloud $P \cup V$ whose richer density lets a LiDAR-based 3D single-object tracker regress more accurate boxes. On nuScenes the augmented input yields mean success/precision of 61.20/73.22 versus 59.22/71.19 for the LiDAR-only baseline, with the largest relative gains on pedestrians and at ranges beyond 30 meters. The paper further claims the virtual cues are general: attaching them to a representative motion-based tracker improves its car success/precision by 1.93/2.16 and its pedestrian success/precision by 4.54/4.36.

Load-bearing premise

The gains depend on having accurate 2D instance masks for the target, but the paper never states whether those masks come from ground-truth annotations or a learned detector, and the depth proxy in Eq. 5 is a second simplification that can misplace boundary cues.

Editorial extensions

If this is right

  • LiDAR-only 3D trackers can consume the augmented point cloud without architectural changes; the paper demonstrates this by improving a second motion-based tracker with the same virtual cues.
  • Small and distant objects, the hardest cases for sparse LiDAR, receive the largest benefit: at ranges of 30 meters or more, pedestrian success rises from 34.14 to 49.14 compared with a motion-based tracker.
  • The approach remains real-time, running at 32.1 FPS on a single GPU while a multimodal feature-fusion tracker runs at 13.2 FPS, so the virtual-cue route avoids the speed cost of dense feature fusion.
  • Degrading image resolution to 800x450 still yields a car tracking success of 64.53, indicating resilience to lower-quality 2D segmentation inputs.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: the source of the 2D instance masks is not specified; if the experiments use ground-truth masks rather than a learned segmentor's outputs, the reported gains may shrink when the masks are noisy.
  • Beyond the paper: the nearest-neighbor depth assignment in Eq. 5 tends to give virtual cues the depth of the closest LiDAR point, so cues near object boundaries can inherit background depth; a learned depth completion module could be a testable extension that yields larger gains.
  • Beyond the paper: because the augmentation is purely input-level, the same virtual-cue recipe should transfer to voxel- and pillar-based trackers; the paper only demonstrates it on point-based motion trackers, leaving that transfer as a concrete open test.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes MVCP (Multimodal-guided Virtual Cues Projection), a scheme that generates dense 3D virtual cues from 2D instance segmentation masks and LiDAR depth, and integrates these cues into a 3D single-object tracking framework called MVCTrack. The method augments raw LiDAR point clouds with virtual points, feeds the augmented cloud into a LiDAR-based tracker (a reimplementation of the P2P baseline [22]), and reports gains on the nuScenes dataset, including mean success/precision of 61.20/73.22 versus 59.22/71.19 for the baseline. The paper also reports a generalization experiment where virtual cues improve another tracker, M2Track, and an ablation of sampling strategies.

Significance. If the reported gains are reproducible and the virtual cues come from a deployable 2D segmentor, the proposed scheme would be a practical, lightweight way to improve LiDAR-based 3D tracking in sparse scenarios, with the plug-and-play property being a genuine strength. The generalization experiment in Table VI, showing gains when applying MVCP to M2Track, provides independent evidence that the mechanism is not tied to one tracker. The paper is also honest about the depth-association limitation, noting that more advanced depth completion is future work. However, the validation is weakened by an undisclosed mask source, a copy-paste error in Table III, and a post-hoc choice of sampling strategy, so the central claim is plausible but not yet rigorously established.

major comments (4)
  1. [Section IV-B, Eqs. (3)-(5), Table I] The source of the 2D instance masks Bob_j is never specified. The text says only that they come 'from an 2D segmentor' without identifying the architecture, weights, or whether the masks are predicted at inference or obtained from nuScenes ground-truth annotations. Since the central claim is that image-derived virtual cues from a lightweight 2D detector improve 3D SOT, the reported 1.98/2.03 mean gain over the baseline in Table I is only meaningful if the masks are not oracle annotations. Please specify the segmentor, its training data, and whether masks are predicted or taken from annotations; if the masks are oracle, add an experiment using outputs of an actual detector, including cases with missing masks and false positives.
  2. [Table III] The 'Improvement' row in Table III reports identical values (↑14.53 / ↑9.73) for both Pedestrian and Bicycle, which is a clear copy-paste error. As printed, the table does not support the claimed gains for bicycles. The baseline for the improvement is also ambiguous: it should state whether the comparison is against M2Track, MMFTrack, or another method.
  3. [Section V-B, 'Robustness of 2D Segmentation Quality', Table V] The experiment that reduces image resolution does not actually simulate inaccurate segmentation. Lowering the resolution of the input image while still using the same mask source does not introduce the types of errors expected from a learned segmentor, such as missing masks, false positives, or boundary inaccuracies. As stated, the experiment cannot support the conclusion that MVCTrack is resilient to degraded 2D segmentation quality in a deployment scenario.
  4. [Section V-B, Ablation study of sampling strategy, Table II] The choice of Strategy 3 as the best sampling strategy is made after evaluating all three strategies on the same nuScenes benchmark used for the main results. This post-hoc selection on the test set means the reported gains for Strategy 3 may be optimistic. The paper should report variance across multiple runs or use a validation split to select the strategy before reporting test performance.
minor comments (5)
  1. [Throughout] There are numerous typos and grammatical errors, including 'bagins' (Section IV-B), 'from an 2D segmentor' (Section IV-B), 'the the baseline' (Section V-B), and 'Mutimodal' in Figure 1. A careful language edit is needed.
  2. [Section IV-B, Eq. (6)] Equation (6) appears to define the unprojection of the virtual cue but the notation mixes coordinates and depth; please clarify the relationship between (xi, yi) in image space and (xi, yi, zi) in 3D space, and ensure the depth variable is consistently named.
  3. [Section V-B, Table I] Several rows in Table I have missing entries (e.g., V2B for Truck/Trailer/Bus, SMAT for Pedestrian/Truck/Trailer, FlowTrack for Truck/Trailer, MMFTrack for Pedestrian/Truck/Trailer). If these methods were not evaluated on those categories, state that explicitly; otherwise, fill in the values or mark them clearly as not reported.
  4. [Section V-A] The baseline [22] is cited as an arXiv preprint and is also the backbone of MVCTrack. Please clarify whether the baseline numbers are from the official paper or from the authors' reimplementation, and provide the version/date of the arXiv preprint to aid reproducibility.
  5. [Section V-B, Running Speed] The running speed is reported only as a single number (32.1 FPS) without details on hardware settings beyond the GPU model; please specify whether this includes the 2D segmentation time, and report the speed of the baseline for a fair comparison.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the virtual-cue gains are tested on an independent tracker (M2Track), and the main remaining concern is the undisclosed 2D mask source, which is an input-validity issue rather than a circularity.

full rationale

The paper's load-bearing claim is that image-derived virtual cues improve 3D SOT. This is established empirically: the augmented point cloud Paug = P ∪ V is fed into a trained tracker, and Table I reports gains over a reimplemented baseline [22], while Table VI transfers the same cues to M2Track and obtains 1.93/2.16 and 4.54/4.36 gains in Car/Pedestrian. The M2Track experiment is independent of the authors' own tracker and shows the cues themselves help, so the central result does not reduce to a fitted parameter or a self-citation. The only mild concern is that the main baseline [22] shares authors (S. Zhou, Z. Yuan) with the present paper and the comparison is reported as the headline improvement; this is a self-citation in the evaluation, but it is not load-bearing because the M2Track generalization result provides outside evidence. The sampling-strategy selection in Table II is a standard choice of hyperparameter on the validation benchmark and is not a prediction from a fitted quantity. The paper does not invoke a uniqueness theorem, and it does not rename a known result: it explicitly cites MVP [18] for the virtual-point idea. The undisclosed provenance of the 2D instance masks (oracle vs. predicted) is a real validity risk for real-world deployment, but it is an input-specification gap, not a circular derivation; Eq. (5)'s nearest-neighbor depth proxy is acknowledged as future work and does not make the output equivalent to the input by construction. Overall, the derivation chain is empirical and self-contained; score 2 reflects the minor self-cited baseline and benchmark-level hyperparameter selection, not circularity.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The method rests on three unproven assumptions: availability of accurate 2D instance masks, validity of nearest-neighbor depth in image space, and known sensor calibration. No new physical entities are introduced; the virtual cues are synthetic data points created from the raw input.

free parameters (2)
  • tau (virtual cues per object) = not reported
    Eq (3)-(4) sample a fixed number tau of virtual cues per detection. The value is never given, and no ablation varies it, though Table II varies the sampling strategy.
  • sampling strategy = Strategy 3: fixed tau for all objects
    The final method adopts Strategy 3 after comparing strategies on the same dataset (Table II), so the reported performance is partly selected from the test distribution.
assumptions (3)
  • domain assumption 2D instance masks for objects are available at every keyframe
    The MVCP pipeline begins with 'a set of 2D object segmentation masks Bob_j from an 2D segmentor' (Section IV.B). The paper does not validate with a real detector and may rely on oracle masks from nuScenes.
  • ad hoc to paper Virtual cue depth equals the nearest projected LiDAR point's depth
    Eq (3) and Eq (5) set zi and then choose di from the nearest neighbor in image space. This creates correct depths only where the LiDAR depth is locally constant; boundaries and occlusions are distorted. The paper calls this a limitation and leaves it to future work.
  • domain assumption Accurate LiDAR-camera calibration and motion compensation
    Eq (1)-(2) require extrinsic and intrinsic matrices and temporal alignment, which are taken for granted from the nuScenes dataset.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MVCTrack: Boosting 3D Point Cloud Tracking via Multimodal-Guided Virtual Cues." pith.science (2026). https://pith.science/paper/AS3UZXIO

@misc{pith2026241202734,
  author       = {Pith},
  title        = {Pith review of: MVCTrack: Boosting 3D Point Cloud Tracking via Multimodal-Guided Virtual Cues},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AS3UZXIO}},
  note         = {Machine review of arXiv:2412.02734}
}
read the original abstract

3D single object tracking is essential in autonomous driving and robotics. Existing methods often struggle with sparse and incomplete point cloud scenarios. To address these limitations, we propose a Multimodal-guided Virtual Cues Projection (MVCP) scheme that generates virtual cues to enrich sparse point clouds. Additionally, we introduce an enhanced tracker MVCTrack based on the generated virtual cues. Specifically, the MVCP scheme seamlessly integrates RGB sensors into LiDAR-based systems, leveraging a set of 2D detections to create dense 3D virtual cues that significantly improve the sparsity of point clouds. These virtual cues can naturally integrate with existing LiDAR-based 3D trackers, yielding substantial performance gains. Extensive experiments demonstrate that our method achieves competitive performance on the NuScenes dataset.

Figures

Figures reproduced from arXiv: 2412.02734 by the authors.

Figure 1
Figure 1. (a) Left: statistics of the number of points on nuScenes’s car. Right: the number of points on nuScenes’s car with multimodal-guided virtual cues. Raw points is red, the virtual cues is blue. (b) Virtual cues generated using MVCP scheme. Blue Square box: Virtual cues of a certain frame. The raw points are marked in gray, and the virtual cues are marked in blue. search space, thereby improving tracking accuracy. Howe… view at source ↗
Figure 2
Figure 2. MVCTrack Framework. Firstly, Multimodal-guided virtual cues projection, this step will generate virtual cues based [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Visualization results on nuScenes dataset. We compare our MVCTrack with M [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 35 canonical work pages

  1. [22]

    P2p: Part- to-part motion cues guide a strong tracking framework for lidar point clouds,

    J. Nie, F. Xie, S. Zhou, X. Zhou, D.-K. Chae, and Z. He, “P2p: Part- to-part motion cues guide a strong tracking framework for lidar point clouds,” arXiv e-prints, pp. arXiv–2407, 2024

  2. [1]

    Objects as points,

    X. Zhou, D. Wang, and P. Kr ¨ahenb¨uhl, “Objects as points,” in arXiv preprint arXiv:1904.07850, 2019

  3. [2]

    Ptt: Point-track-transformer module for 3d single object tracking in point clouds,

    J. Shan, S. Zhou, Z. Fang, and Y . Cui, “Ptt: Point-track-transformer module for 3d single object tracking in point clouds,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2021, pp. 1310–1316

  4. [3]

    Fastpillars: A deployment-friendly pillar-based 3d detector

    S. Zhou, Z. Tian, X. Chu, X. Zhang, B. Zhang, X. Lu, C. Feng, and Z. Jie, “Fastpillars: A deployment-friendly pillar-based 3d detector.”

  5. [4]

    Eagermot: 3d multi-object track- ing via sensor fusion,

    A. Kim, A. O ˇsep, and L. Leal-Taix´e, “Eagermot: 3d multi-object track- ing via sensor fusion,” in Proceedings of the 2021 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2021, pp. 11 315–11 321

  6. [5]

    Real-time 3d single object tracking with transformer,

    J. Shan et al., “Real-time 3d single object tracking with transformer,” IEEE Transactions on Multimedia , vol. 25, pp. 2339–2353, 2022

  7. [6]

    Point siamese network for person tracking using 3d point clouds,

    Y . Cui, Z. Fang, and S. Zhou, “Point siamese network for person tracking using 3d point clouds,” Sensors, vol. 20, no. 1, p. 143, 2019

  8. [7]

    P2b: Point-to-box network for 3d object tracking in point clouds,

    H. Qi, C. Feng, Z. Cao, F. Zhao, and Y . Xiao, “P2b: Point-to-box network for 3d object tracking in point clouds,” in computer vision and pattern recognition , 2020, pp. 6329–6338

Show all 42 references
  1. [8]

    3d-siamrpn: An end-to- end learning method for real-time 3d single object tracking using raw point cloud,

    Z. Fang, S. Zhou, Y . Cui, and S. Scherer, “3d-siamrpn: An end-to- end learning method for real-time 3d single object tracking using raw point cloud,” IEEE Sensors Journal , vol. 21, no. 4, pp. 4995–5011, 2020

  2. [9]

    Real-time 3d single object tracking with transformer,

    J. Shan, S. Zhou, Y . Cui, and Z. Fang, “Real-time 3d single object tracking with transformer,” IEEE Transactions on Multimedia, vol. 25, pp. 2339–2353, 2022

  3. [10]

    3d object tracking with transformer,

    Y . Cui, Z. Fang, J. Shan, Z. Gu, and S. Zhou, “3d object tracking with transformer,” British Machine Vision Conference, pp. 1445–1458, 2021

  4. [11]

    Box-aware feature enhancement for single object tracking on point clouds,

    C. Zheng, X. Yan, J. Gao, W. Zhao, W. Zhang, Z. Li, and S. Cui, “Box-aware feature enhancement for single object tracking on point clouds,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 13 199–13 208

  5. [12]

    3d siamese voxel-to- bev tracker for sparse point clouds,

    L. Hui, L. Wang, M. Cheng, J. Xie, and J. Yang, “3d siamese voxel-to- bev tracker for sparse point clouds,” Advances in Neural Information Processing Systems, vol. 34, pp. 28 714–28 727, 2021

  6. [13]

    Fully-convolutional siamese networks for object tracking,

    L. Bertinetto, J. Valmadre, J. F. Henriques, A. Vedaldi, and P. H. Torr, “Fully-convolutional siamese networks for object tracking,” in Com- puter Vision–ECCV 2016 Workshops: Amsterdam, The Netherlands, October 8-10 and 15-16, 2016, Proceedings, Part II 14 . Springer, 2016, pp...

  7. [14]

    High performance visual tracking with siamese region proposal network,

    B. Li, J. Yan, W. Wu, Z. Zhu, and X. Hu, “High performance visual tracking with siamese region proposal network,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 8971–8980

  8. [15]

    Learning localization- aware target confidence for siamese visual tracking,

    J. Nie, Z. He, Y . Yang, M. Gao, and Z. Dong, “Learning localization- aware target confidence for siamese visual tracking,” IEEE Transac- tions on Multimedia , 2022

  9. [16]

    F-siamese tracker: A frustum-based double siamese network for 3d single object tracking,

    H. Zou, J. Cui, X. Kong, C. Zhang, Y . Liu, F. Wen, and W. Li, “F-siamese tracker: A frustum-based double siamese network for 3d single object tracking,” in IEEE/RSJ International Conference on Intelligent Robots and Systems , 2020, pp. 8133–8139

  10. [17]

    Mmf-track: Multi-modal multi-level fusion for 3d single object tracking,

    Z. Li et al., “Mmf-track: Multi-modal multi-level fusion for 3d single object tracking,” IEEE Transactions on Intelligent Vehicles , 2023

  11. [18]

    Multimodal virtual point 3d detection,

    T. Yin, X. Zhou, and P. Kr ¨ahenb¨uhl, “Multimodal virtual point 3d detection,” in Advances in Neural Information Processing Systems , vol. 34, 2021, pp. 16 494–16 507

  12. [19]

    Yolov6: A single-stage object detection framework for industrial applications,

    C. Li et al. , “Yolov6: A single-stage object detection framework for industrial applications,” arXiv preprint arXiv:2209.02976 , 2022

  13. [20]

    Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors,

    C.-Y . Wang, A. Bochkovskiy, and H.-Y . M. Liao, “Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 7464–7475

  14. [21]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Beijbom, “nuscenes: A multimodal dataset for autonomous driving,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 11 621–11 631

  15. [23]

    Lever- aging shape completion for 3d siamese tracking,

    S. Giancola, J. Zarzar, B. Ghanem, S. Giancola, and J. Zarzar, “Lever- aging shape completion for 3d siamese tracking,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 1359–1368

  16. [24]

    Deep hough voting for 3d object detection in point clouds,

    C. R. Qi, O. Litany, K. He, and L. J. Guibas, “Deep hough voting for 3d object detection in point clouds,” in proceedings of the IEEE/CVF International Conference on Computer Vision , 2019, pp. 9277–9286

  17. [25]

    3d siamese transformer network for single object tracking on point clouds,

    L. Hui, L. Wang, L. Tang, K. Lan, J. Xie, and J. Yang, “3d siamese transformer network for single object tracking on point clouds,” in Computer Vision–ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part II . Springer, 2022, pp. 293–310

  18. [26]

    Glt-t: Global-local transformer voting for 3d single object tracking in point clouds,

    J. Nie, Z. He, Y . Yang, M. Gao, and J. Zhang, “Glt-t: Global-local transformer voting for 3d single object tracking in point clouds,” in Proceedings of the AAAI Conference on Artificial Intelligence , 2023, pp. 1957–1965

  19. [27]

    Pttr: Relational 3d point cloud object tracking with transformer,

    C. Zhou, Z. Luo, Y . Luo, T. Liu, L. Pan, Z. Cai, H. Zhao, and S. Lu, “Pttr: Relational 3d point cloud object tracking with transformer,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 8531–8540

  20. [28]

    Cmt: Context- matching-guided transformer for 3d tracking in point clouds,

    Z. Guo, Y . Mao, W. Zhou, M. Wang, and H. Li, “Cmt: Context- matching-guided transformer for 3d tracking in point clouds,” in Computer Vision–ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part XXII. Springer, 2022, pp. 95–111

  21. [29]

    A lightweight and detector-free 3d single object tracker on point clouds,

    Y . Xia, Q. Wu, W. Li, A. B. Chan, and U. Stilla, “A lightweight and detector-free 3d single object tracker on point clouds,” IEEE Transactions on Intelligent Transportation Systems , 2023

  22. [30]

    Osp2b: One- stage point-to-box network for 3d siamese tracking,

    J. Nie, Z. He, Y . Yang, Z. Bao, M. Gao, and J. Zhang, “Osp2b: One- stage point-to-box network for 3d siamese tracking,” in Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, 2023, pp. 1285–1293

  23. [31]

    Temporal-aware siamese tracker: Integrate temporal context for 3d object tracking,

    K. Lan, H. Jiang, and J. Xie, “Temporal-aware siamese tracker: Integrate temporal context for 3d object tracking,” in Proceedings of the Asian Conference on Computer Vision , 2022, pp. 399–414

  24. [32]

    Synchronize feature extracting and matching: A single branch framework for 3d object tracking,

    T. Ma, M. Wang, J. Xiao, H. Wu, and Y . Liu, “Synchronize feature extracting and matching: A single branch framework for 3d object tracking,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 9953–9963

  25. [33]

    Pillartrack: Redesigning pillar-based transformer network for single object tracking on point clouds,

    W. Xu, S. Zhou, and Z. Yuan, “Pillartrack: Redesigning pillar-based transformer network for single object tracking on point clouds,” arXiv preprint arXiv:2404.07495, 2024

  26. [34]

    Beyond 3d siamese tracking: A motion-centric paradigm for 3d single object tracking in point clouds,

    C. Zheng, X. Yan, H. Zhang, B. Wang, S. Cheng, S. Cui, and Z. Li, “Beyond 3d siamese tracking: A motion-centric paradigm for 3d single object tracking in point clouds,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 8111–8120

  27. [35]

    An effective motion- centric paradigm for 3d single object tracking in point clouds,

    C. Zheng, X. Yan, H. Zhang, S. Cui, and Z. Li, “An effective motion- centric paradigm for 3d single object tracking in point clouds,” IEEE Transactions on Pattern Analysis and Machine Intelligence , 2023

  28. [36]

    Exploiting more information in sparse point cloud for 3d single object tracking,

    Y . Cui, J. Shan, Z. Gu, Z. Li, and Z. Fang, “Exploiting more information in sparse point cloud for 3d single object tracking,” IEEE Robotics and Automation Letters , vol. 7, no. 4, pp. 11 926–11 933, 2022

  29. [37]

    Sttracker: Spatio-temporal tracker for 3d single object tracking,

    Y . Cui, Z. Li, and Z. Fang, “Sttracker: Spatio-temporal tracker for 3d single object tracking,” IEEE Robotics and Automation Letters , 2023

  30. [38]

    Towards category unification of 3d single object tracking on point clouds,

    J. Nie, Z. He, X. Lv, X. Zhou, D.-K. Chae, and F. Xie, “Towards category unification of 3d single object tracking on point clouds,” in The Twelfth International Conference on Learning Representations , 2024

  31. [39]

    Flowtrack: Point- level flow network for 3d single object tracking,

    S. Li, Y . Cui, L. Zhiheng, and Z. Fang, “Flowtrack: Point- level flow network for 3d single object tracking,” arXiv preprint arXiv:2407.01959, 2024

  32. [40]

    Mbptrack: Improv- ing 3d point cloud tracking with memory networks and box priors,

    T.-X. Xu, Y .-C. Guo, Y .-K. Lai, and S.-H. Zhang, “Mbptrack: Improv- ing 3d point cloud tracking with memory networks and box priors,” arXiv preprint arXiv:2303.05071 , 2023

  33. [41]

    Online object tracking: A bench- mark,

    Y . Wu, J. Lim, and M.-H. Yang, “Online object tracking: A bench- mark,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2013, pp. 2411–2418

  34. [42]

    A novel performance evaluation methodology for single-target trackers,

    M. Kristan, J. Matas, A. Leonardis, T. V oj ´ıˇr, R. Pflugfelder, G. Fer- nandez, G. Nebehay, F. Porikli, and L. ˇCehovin, “A novel performance evaluation methodology for single-target trackers,” IEEE transactions on pattern analysis and machine intelligence , vol. 38, no. 11,...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.