Pith. sign in

REVIEW 4 major objections 5 minor 84 references

Video Individual Counting and Tracking from Moving Drones: A Benchmark and Methods

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Moving-drone video counting and tracking works by matching pixel-level pedestrian descriptors instead of locating each person.

desk verdict Worth a serious referee: the dataset is a real contribution and the method ideas are plausible, but the evaluation tables are internally inconsistent enough that the headline gains can't be trusted until the numbers are corrected. read the letter →

arxiv 2601.12500 v2 pith:KUJPN3NU submitted 2026-01-18 cs.CV

Yaowu Fan , Jia Wan , Tao Han , Andy J. Ma , Wanli Ouyang , Antoni B. Chan This is my paper
classification cs.CV
keywords videoindividualcountingdensecrowddronetrackingoptimaltransportdescriptormatchingadaptivedustbinscoredensitymapdecompositionmovingdatasetmulti-object
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 claims that counting and tracking unique pedestrians in dense crowds filmed from moving drones should be done by matching pixel-level visual descriptors of heads across frames, not by detecting and localizing each individual. It introduces MovingDroneCrowd++, a large drone-video dataset with dense crowds, varied altitudes, angles, and lighting, and two methods built on this idea: GD3A counts by decomposing a global density map into shared, inflow, and outflow maps, while DVTrack turns descriptor matches into instance tracks by voting. On this dataset, the methods cut counting error by 47.4% and raise tracking HOTA by 39.2% over previous approaches. If correct, the core insight is that explicit pedestrian localization, which is fragile in dense and fast-moving aerial scenes, can be replaced by correspondence between many small features, making counting and tracking more robust.

What carries the argument

The central mechanism is optimal transport on pixel-level descriptors with an input-adaptive dustbin score. A backbone plus pre-trained density estimator filters feature maps to head regions; each head yields multiple descriptors. These are position-enhanced and refined by an attentional graph neural network, then matched across frames by solving a Sinkhorn-regularized transport problem whose cost matrix includes a learnable dustbin score predicted from the current descriptor pair (rather than a fixed dataset-level constant). The matching matrix decomposes global density maps into shared/inflow/outflow components, and a reverse top-k strategy plus Hungarian assignment over descriptor votes y

What would settle it

A direct test would be to annotate true pixel-level correspondences for a few dozen head pairs in the MovingDroneCrowd++ test set (e.g., by manual keypoint matching) and compare them against the matches predicted by GD3A; if the descriptor-matching accuracy on those verified pairs is near chance for heads that visibly change scale or rotate, the training-signal assumption is violated and the claimed mechanism would be suspect.

Watch

Extended reading notes

Core claim

The authors propose that pixel-level optimal-transport matching of pedestrian head descriptors, with a per-frame-pair adaptive dustbin score, is enough to drive accurate video-level individual counting and tracking, without ever explicitly localizing or detecting pedestrians. GD3A uses these correspondences to split each frame's global density map into shared, inflow, and outflow density maps, summing inflow to get the unique-person count; DVTrack reuses the same matches to build a voting matrix that assigns pedestrians across frames. The central claim is that intra-group matching errors among descriptors of the same pedestrian do not hurt the final result, and that pixel-level association a

Load-bearing premise

The pixel-level supervision assumes that a pedestrian head translates rigidly between nearby frames (p + Δ maps to p' + Δ for all small offsets), but a moving drone also changes scale, rotation, and perspective, so those pseudo-labels are not exactly correct.

Editorial extensions

If this is right

  • If the descriptor-matching approach generalizes, video individual counting in dense crowds no longer depends on accurate head localization, which typically degrades in crowded low-altitude aerial views.
  • A single learned association module can serve both counting and tracking, since DVTrack derives trajectories from the same pixel-level correspondences without extra training.
  • The adaptive dustbin score implies performance should hold across a wide range of frame intervals and drone speeds, as the paper's interval-sweep experiment shows.
  • Because the method is density-map based and avoids cross-frame cross-attention, it should be computationally cheaper than prior VIC methods while being more interpretable, since inflow/outflow maps are explicit.
  • The dataset, being the largest moving-drone dense-crowd video benchmark with full trajectory annotations, should set a new evaluation standard for both counting and tracking under aerial motion.

Reading between the lines

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

  • The rigid-translation pseudo-label (Eq. 17) used to supervise pixel-level matches is only an approximation: under a moving drone, head motion between frames includes scale, rotation, and perspective changes, so the training signal may inject systematic errors that the paper does not measure.
  • A natural extension is to test whether the same pipeline works on other aerial platforms (e.g., fixed-wing or high-altitude) or on crowd sub-tasks like density-based flow estimation, where the inflow/outflow decomposition could be reused.
  • Because the method avoids explicit localization, it may also be transferable to dense counting of other small objects (e.g., animals or vehicles) in aerial video, if a descriptor-level matching signal can be constructed.
  • The claimed robustness to density-map error suggests a possible soft-supervision variant that trains the counter jointly with the matcher, rather than using a frozen pre-trained counter, potentially improving performance further.
Share X Bluesky LinkedIn Reddit HN

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 introduces MovingDroneCrowd++, a dataset of 120 video clips captured by moving drones in dense, large-scale crowd scenes with full head-bounding-box and identity annotations, and two methods: GD3A for video individual counting and DVTrack for multi-object tracking. GD3A decomposes a predicted global density map into shared, inflow, and outflow density maps using pixel-level pedestrian descriptor matching via optimal transport with an adaptive dustbin score. DVTrack converts the same descriptor matches into instance-level associations through descriptor voting. On MovingDroneCrowd++, the authors report a 47.4% reduction in counting MAE and a 39.2% (or, per the abstract, 64.6%) improvement in tracking HOTA over prior methods; experiments on the fixed-camera VSCrowd benchmark also show state-of-the-art counting results. The paper includes ablations on the matching components, the adaptive dustbin score, frame sampling interval, and sensitivity to global density estimation.

Significance. If the reported results are validated, the paper makes a strong contribution: MovingDroneCrowd++ is the largest moving-drone dense-crowd video dataset with full trajectory annotations, and the descriptor-association design is a plausible route to avoiding hard localization in dense crowds. Strengths include the public release of code, dataset, and pretrained models; evaluation on an external fixed-camera benchmark (VSCrowd); and a well-structured ablation program (Tables V–VI, Fig. 7). However, the current evaluation tables contain serious internal inconsistencies, and the pseudo-label model that supervises the descriptor matching (Eq. 17) is unvalidated for the moving-drone regime. These issues affect the central quantitative claims and must be resolved before the results can be considered credible.

major comments (4)
  1. [§V.B, Tables II, III, V, VI] The headline numbers do not agree across tables. GD3A MAE is 40.11 in Table II, 32.25 in Table VI (row 'Ours' with ADS), and 66.61 in Table V (row with all components). Since Table VI is an ablation of the dustbin score and Table V an ablation of the matching process, the fully configured model should appear with consistent MAE; the paper does not specify which subset (test vs validation) or which frame-interval/global-density setting each table uses. Additionally, Table III reports improvement arrows that do not match the tabulated numbers: HOTA 20.9→34.4 is +64.6%, not +39.2%, and AssA 17.6→41.2 is +134%, not +40%. The abstract states a tracking improvement of 64.6% while the introduction and conclusion state 39.2%. These inconsistencies must be reconciled and the reporting settings for each table stated explicitly.
  2. [§IV.D, Eq. (17)] The pseudo-label supervision assumes a rigid translation of head patches: p_t+Δ ↔ p_{t+δ}+Δ for all Δ with ||Δ||_∞ < r. For a moving drone with changing altitude and camera angle, the apparent motion of a head includes scale, rotation, and perspective distortion; the error grows with |Δ| and with the frame interval δ. Training samples δ randomly from 3 to 8 frames, while the test-time evaluation in Fig. 7(a) extends to 6 s, so the low-validity regime is exactly the one the paper claims robustness in. Since Eq. 17 generates the supervision for the descriptor association that drives both GD3A and DVTrack, a systematic label bias could corrupt the density decomposition and the voting-based tracking. The paper provides no validation of this assumption (e.g., an analysis of pseudo-label correctness or a comparison with an affine/homography-based pseudo-label). This is load-bearing and needs a
  3. [§V.C.2, Table VI] DRNet is described as an 'OT-based instance-level matching method' in the ablation, but in §V.B.1 and Table II it is grouped under 'Localization-based VIC Methods'. This inconsistency makes it impossible to interpret the adaptive-dustbin comparison: if DRNet is not an OT-based method, the controlled comparison is not what the text claims. Please clarify the categorization and describe the exact DRNet configuration used in the ablation.
  4. [§V.C.3, Fig. 7(a)] The robustness claim with respect to frame sampling interval is based only on a qualitative plot. No quantitative numbers, error bars, or a description of the subset/scoring protocol are given. Provide the MAE/RMSE values at representative intervals (e.g., 1s, 3s, 6s) for GD3A and the compared methods so the reader can verify the claimed stable performance up to 6 s.
minor comments (5)
  1. [Table II] The column header 'MIOE' should be 'MOAE' per the metric definitions in §V.A.2 (Eq. 19).
  2. [Table IV] There are missing spaces in two cells: '8.525.0' and '24.947.9' should likely be '8.5 25.0' and '24.9 47.9'.
  3. [Abstract / Introduction / Conclusion] The tracking improvement is reported as 64.6% in the abstract but 39.2% in the introduction and conclusion. After reconciling Table III, please align all occurrences.
  4. [Table V] The first metric column is labeled 'MSE', but the paper defines and uses MAE as the counting metric; MSE is not defined. This appears to be a typo and should be corrected.
  5. [§V.B.3] The method is referred to as 'DVTracker' in the text, but the method name used elsewhere is 'DVTrack'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical benchmark and method evaluation with held-out test splits and external validation.

full rationale

The paper's central claims are empirical: GD3A and DVTrack are trained and evaluated on held-out test splits of MovingDroneCrowd++ and on the external VSCrowd benchmark. The reported 47.4% counting-error reduction and 39.2% HOTA improvement are measured against concrete baselines, including the authors' own SDNet, but SDNet is explicitly identified as a baseline and the comparison is arithmetic, not definitional. The method's components—OT-based descriptor matching (Eq. 9), adaptive dustbin scoring (Eq. 11), density-map decomposition, and descriptor voting—are each defined independently of the evaluation targets. The pixel-level supervision in Eq. 17 (p_t+Δ ↔ p_{t+δ}+Δ) is an approximation used to generate training correspondence labels; it is an untested assumption about apparent motion under a moving drone, and therefore a correctness/robustness concern, but it is not circular because the method's outputs are not reinserted as inputs or definitions. The dataset claim ('largest ... captured by moving drones') is supported by the dataset statistics and Table I rather than by a self-citation chain. No load-bearing uniqueness theorem, ansatz, or fitted parameter is imported from prior work and then declared a prediction. Overall, the derivation chain is self-contained as an empirical study, and no step reduces by construction to its own inputs.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The central claim depends on several hand-chosen hyperparameters that are not reported, and on a set of domain assumptions about head-patch motion, density-maxima localization, and the representativeness of internet-sourced drone video. No new physical or conceptual entities are introduced.

free parameters (7)
  • filter threshold tau
    Predefined threshold in Eq. 4 to keep head descriptors; value not reported.
  • top-K
    Parameter in the reverse top-k association strategy, Eq. 14; value not reported.
  • matching threshold theta
    Threshold in Eq. 15 to accept a descriptor match; value not reported.
  • local patch radius r
    Radius in Eq. 17 that defines the head region copied from point annotations to pixel labels; value not reported.
  • AGNN depth L and dimensions
    Depth and feature dimensions of the attentional GNN in Eq. 6-8; not specified.
  • Sinkhorn iterations and entropic regularization
    Details of solving Eq. 9 via Sinkhorn (Cuturi) not given.
  • pre-trained image-level counter
    Used to generate global density maps in Eq. 3; its training set and architecture are not specified, only that it is pre-trained.
assumptions (5)
  • domain assumption Local translational motion of head patches (Eq. 17)
    Pixel-level correspondence labels are generated by copying the same local offset between frames; this assumes no scale, rotation, or perspective deformation of heads between frames.
  • domain assumption Density map local maxima correspond to individual pedestrian heads
    DVTrack and descriptor coordinate assignment use local maxima of the global density map as pedestrian locations; in very dense crowds this may merge or miss heads.
  • standard math The global density map sums to pedestrian count
    Standard density-map crowd counting assumption; also used in Eq. 1.
  • domain assumption Internet-sourced videos are representative of real-world dense crowd scenes
    The dataset is collected from Bilibili, YouTube, Google, Bing; selection and compression artifacts may limit representativeness.
  • domain assumption Head descriptors are consistent for the same pedestrian across frames
    The whole matching/tracking framework assumes appearance descriptors of the same head are more similar than those of different heads, even under illumination and viewpoint changes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Video Individual Counting and Tracking from Moving Drones: A Benchmark and Methods." pith.science (2026). https://pith.science/paper/KUJPN3NU

@misc{pith2026260112500,
  author       = {Pith},
  title        = {Pith review of: Video Individual Counting and Tracking from Moving Drones: A Benchmark and Methods},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KUJPN3NU}},
  note         = {Machine review of arXiv:2601.12500}
}
read the original abstract

Counting and tracking dense crowds in large-scale scenes is a highly practical yet challenging problem. Existing methods mostly rely on fixed-camera datasets with limited scene coverage, making them inadequate for crowd analysis in large-scale scenes. To bridge this gap, we introduce MovingDroneCrowd++, the largest video-level dataset dedicated to dense crowd counting and tracking with fast-moving drones, captured under diverse flight altitudes, camera angles, and illumination conditions. Existing methods, however, still fail to achieve satisfactory video individual counting or tracking performance under these challenging aerial conditions. To this end, we propose GD3A (Global Density map Decomposition via group-wise Descriptor Association), a video individual counting method that first establishes pixel-level correspondences between pedestrian descriptors across frames via optimal transport with an adaptive dustbin score. Then, group-wise association is adopted to guide the decomposition of the global density map into shared, inflow, and outflow density maps. We further introduce a pedestrian tracking method, DVTrack (Descriptor Voting Track), which converts descriptor-level matching into instance-level association through descriptor voting. Our methods rely on the association results of group-wise multiple descriptors for each pedestrian rather than a single vector. Since intra-group matching errors do not affect the final counting and tracking results, our methods are more robust in dense crowds and challenging aerial conditions. Experiments show that our methods achieve substantial gains in both crowd counting and tracking on moving-drone videos with dense crowds and complex motions, reducing counting error by 47.4% and improving tracking accuracy by 64.6%. Code, dataset, and pretrained models are available at https://github.com/fyw1999/MovingDroneCrowd.

Figures

Figures reproduced from arXiv: 2601.12500 by the authors.

Figure 1
Figure 1. Comparison between existing crowd analysis datasets and ours. Existing research has predominantly focused on (a) free-viewpoint images captured [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Exemplars from the MovingDroneCrowd++ dataset. Due to space constraints, only two frames are displayed for each video clip. Each frame is [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Crowd density statistics of the MovingDroneCrowd++ dataset. (a) [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Scene attributes statistics of the MovingDroneCrowd++ dataset. (a) [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The pipeline of the proposed GD3A. Given two frames Ft and Ft+δ, a backbone extracts feature maps Ft and Ft+δ, which are filtered using global density maps Dˆ g t and Dˆ g t+δ predicted by a pre-trained estimator to retain visual descriptors for pedestrian heads. Subse…
Figure 6
Figure 6. Figure 6: Visual comparison of inflow and output density maps predicted by our method and other density map-based VIC methods. For each frame pair, [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: (a) Comparison of our method with other methods under different frame intervals, ranging from [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Visual comparisons of tracking results. Other methods suffer from frequent ID switches and localization errors, whereas our method DVTrack [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

84 extracted references · 3 linked inside Pith

  1. [1]

    Integrated sensing and communication for low altitude econ- omy: Opportunities and challenges,

    Y . Jiang, X. Li, G. Zhu, H. Li, J. Deng, K. Han, C. Shen, Q. Shi, and R. Zhang, “Integrated sensing and communication for low altitude econ- omy: Opportunities and challenges,”IEEE Communications Magazine, pp. 1–7, 2025

    2025

  2. [2]

    Unmanned aerial vehicles based low-altitude economy with lifecycle techno-economic-environmental analysis for sustainable and smart cities,

    Y . Zhou, “Unmanned aerial vehicles based low-altitude economy with lifecycle techno-economic-environmental analysis for sustainable and smart cities,”Journal of Cleaner Production, vol. 499, p. 145050, 2025

    2025

  3. [3]

    Low-altitude intelligent transportation: System architecture, infrastructure, and key technologies,

    C. Huang, S. Fang, H. Wu, Y . Wang, and Y . Yang, “Low-altitude intelligent transportation: System architecture, infrastructure, and key technologies,”Journal of Industrial Information Integration, vol. 42, p. 100694, 2024

    2024

  4. [4]

    Learning to count objects in images,

    V . Lempitsky and A. Zisserman, “Learning to count objects in images,” inAdvances in Neural Information Processing Systems, vol. 23. Curran Associates, Inc., 2010

    2010

  5. [5]

    Csrnet: Dilated convolutional neural net- works for understanding the highly congested scenes,

    Y . Li, X. Zhang, and D. Chen, “Csrnet: Dilated convolutional neural net- works for understanding the highly congested scenes,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2018

    2018

  6. [6]

    Regressor- segmenter mutual prompt learning for crowd counting,

    M. Guo, L. Yuan, Z. Yan, B. Chen, Y . Wang, and Q. Ye, “Regressor- segmenter mutual prompt learning for crowd counting,” in2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 28 380–28 389

    2024

  7. [7]

    Rethinking counting and localization in crowds: A purely point-based framework,

    Q. Song, C. Wang, Z. Jiang, Y . Wang, Y . Tai, C. Wang, J. Li, F. Huang, and Y . Wu, “Rethinking counting and localization in crowds: A purely point-based framework,” inProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2021, pp. 3365–3374

    2021

  8. [8]

    Learning from synthetic data for crowd counting in the wild,

    Q. Wang, J. Gao, W. Lin, and Y . Yuan, “Learning from synthetic data for crowd counting in the wild,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2019

    2019

Show all 84 references
  1. [9]

    Multiple object tracking as id prediction,

    R. Gao, J. Qi, and L. Wang, “Multiple object tracking as id prediction,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2025, pp. 27 883–27 893

    2025

  2. [10]

    Drone-assisted public safety net- works: The security aspect,

    D. He, S. Chan, and M. Guizani, “Drone-assisted public safety net- works: The security aspect,”IEEE Communications Magazine, vol. 55, no. 8, pp. 218–223, 2017

    2017

  3. [11]

    Nwpu-crowd: A large-scale benchmark for crowd counting and localization,

    Q. Wang, J. Gao, W. Lin, and X. Li, “Nwpu-crowd: A large-scale benchmark for crowd counting and localization,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020

    2020

  4. [12]

    Composition loss for counting, density map estimation and localization in dense crowds,

    H. Idrees, M. Tayyab, K. Athrey, D. Zhang, S. Al-Maadeed, N. Rajpoot, and M. Shah, “Composition loss for counting, density map estimation and localization in dense crowds,” inProceedings of the European conference on computer vision (ECCV), 2018, pp. 532–546

    2018

  5. [13]

    Single-image crowd counting via multi-column convolutional neural network,

    Y . Zhang, D. Zhou, S. Chen, S. Gao, and Y . Ma, “Single-image crowd counting via multi-column convolutional neural network,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016

    2016

  6. [14]

    Multi-source multi-scale counting in extremely dense crowd images,

    H. Idrees, I. Saleemi, C. Seibert, and M. Shah, “Multi-source multi-scale counting in extremely dense crowd images,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2013

    2013

  7. [15]

    Reducing spatial labeling redundancy for active semi-supervised crowd counting,

    Y . Liu, S. Ren, L. Chai, H. Wu, D. Xu, J. Qin, and S. He, “Reducing spatial labeling redundancy for active semi-supervised crowd counting,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 7, pp. 9248–9255, 2023

    2023

  8. [16]

    Jhu-crowd++: Large-scale crowd counting dataset and a benchmark method,

    V . A. Sindagi, R. Yasarla, and V . M. Patel, “Jhu-crowd++: Large-scale crowd counting dataset and a benchmark method,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 5, pp. 2594– 2609, 2022

    2022

  9. [17]

    Point-to-region loss for semi-supervised point-based crowd counting,

    W. Lin, C. Zhao, and A. B. Chan, “Point-to-region loss for semi-supervised point-based crowd counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2025, pp. 29 363–29 373

    2025

  10. [18]

    Video crowd localization with multifocus gaussian neighborhood at- tention and a large-scale benchmark,

    H. Li, L. Liu, K. Yang, S. Liu, J. Gao, B. Zhao, R. Zhang, and J. Hou, “Video crowd localization with multifocus gaussian neighborhood at- tention and a large-scale benchmark,”IEEE Transactions on Image Processing, vol. 31, pp. 6032–6047, 2022

    2022

  11. [19]

    Scene-adaptive unsupervised crowd counting for video surveillance,

    R. Ma, Y . Hou, C. Li, H. Jia, and X. Xie, “Scene-adaptive unsupervised crowd counting for video surveillance,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 35, no. 7, pp. 6910–6925, 2025

    2025

  12. [20]

    Crowdcl: Unsupervised crowd counting network via contrastive learning,

    Y . Hu, Y . Liu, G. Cao, and J. Wang, “Crowdcl: Unsupervised crowd counting network via contrastive learning,”IEEE Internet of Things Journal, vol. 12, no. 12, pp. 21 704–21 719, 2025

    2025

  13. [21]

    Dancetrack: Multi-object tracking in uniform appearance and diverse motion,

    P. Sun, J. Cao, Y . Jiang, Z. Yuan, S. Bai, K. Kitani, and P. Luo, “Dancetrack: Multi-object tracking in uniform appearance and diverse motion,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2022, pp. 20 993–21 002

    2022

  14. [22]

    Mot20: A bench- mark for multi object tracking in crowded scenes,

    P. Dendorfer, H. Rezatofighi, A. Milan, J. Shi, D. Cremers, I. Reid, S. Roth, K. Schindler, and L. Leal-Taix ´e, “Mot20: A bench- mark for multi object tracking in crowded scenes,”arXiv preprint arXiv:2003.09003, 2020

    2003 arXiv

  15. [23]

    Sportsmot: A large multi-object tracking dataset in multiple sports scenes,

    Y . Cui, C. Zeng, X. Zhao, Y . Yang, G. Wu, and L. Wang, “Sportsmot: A large multi-object tracking dataset in multiple sports scenes,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2023, pp. 9921–9931

    2023

  16. [24]

    Crowded video individual counting informed by social grouping and spatial-temporal displacement priors,

    H. Lu, X. Zhu, W. Zhang, Y . Li, and X. Bai, “Crowded video individual counting informed by social grouping and spatial-temporal displacement priors,”arXiv preprint arXiv:2601.01192, 2026

    2026

  17. [25]

    Detection, tracking, and counting meets drones in crowds: A benchmark,

    L. Wen, D. Du, P. Zhu, Q. Hu, Q. Wang, L. Bo, and S. Lyu, “Detection, tracking, and counting meets drones in crowds: A benchmark,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2021, pp. 7812–7821

    2021

  18. [26]

    Visdrone-cc2021: The vision meets drone crowd counting challenge results,

    Z. Liu, Z. He, L. Wang, W. Wang, Y . Yuan, D. Zhang, J. Zhang, P. Zhu, L. V . Gool, J. Han, S. Hoi, Q. Hu, M. Liu, J. Pan, B. Yin, B. Zhang, C. Liu, D. Ding, D. Liang, G. Ding, H. Lu, H. Lin, J. Chen, J. Li, L. Liu, L. Zhou, M. Shi, Q. Yang, Q. He, S. Peng, W. Xu, W. Han, X. B...

    2021

  19. [27]

    Weakly supervised video individual counting,

    X. Liu, G. Li, Y . Qi, Z. Yan, Z. Han, A. van den Hengel, M.-H. Yang, and Q. Huang, “Weakly supervised video individual counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 19 228–19 237

    2024

  20. [28]

    Detection and tracking meet drones challenge,

    P. Zhu, L. Wen, D. Du, X. Bian, H. Fan, Q. Hu, and H. Ling, “Detection and tracking meet drones challenge,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 11, pp. 7380–7399, 2021

    2021

  21. [29]

    Bytetrack: Multi-object tracking by associating every detection box,

    Y . Zhang, P. Sun, Y . Jiang, D. Yu, F. Weng, Z. Yuan, P. Luo, W. Liu, and X. Wang, “Bytetrack: Multi-object tracking by associating every detection box,” inComputer Vision – ECCV 2022, 2022, pp. 1–21

    2022

  22. [30]

    Track- former: Multi-object tracking with transformers,

    T. Meinhardt, A. Kirillov, L. Leal-Taix ´e, and C. Feichtenhofer, “Track- former: Multi-object tracking with transformers,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2022, pp. 8844–8854. IEEE TRANSACTIONS 15

    2022

  23. [31]

    Sparsetrack: Multi- object tracking by performing scene decomposition based on pseudo- depth,

    Z. Liu, X. Wang, C. Wang, W. Liu, and X. Bai, “Sparsetrack: Multi- object tracking by performing scene decomposition based on pseudo- depth,”IEEE Transactions on Circuits and Systems for Video Technol- ogy, vol. 35, no. 5, pp. 4870–4882, 2025

    2025

  24. [32]

    Multi-object tracking in the dark,

    X. Wang, K. Ma, Q. Liu, Y . Zou, and Y . Fu, “Multi-object tracking in the dark,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 382–392

    2024

  25. [33]

    Diffusiontrack: Diffusion model for multi-object tracking,

    R. Luo, Z. Song, L. Ma, J. Wei, W. Yang, and M. Yang, “Diffusiontrack: Diffusion model for multi-object tracking,”Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 5, pp. 3991–3999, 2024

    2024

  26. [34]

    Dr.vic: Decompo- sition and reasoning for video individual counting,

    T. Han, L. Bai, J. Gao, Q. Wang, and W. Ouyang, “Dr.vic: Decompo- sition and reasoning for video individual counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2022, pp. 3083–3092

    2022

  27. [35]

    Prototype-guided dual- transformer reasoning for video individual counting,

    R. Li, Y . Liu, H. Li, J. Li, and G. Lu, “Prototype-guided dual- transformer reasoning for video individual counting,” 2024, p. 10258–10267

    2024

  28. [36]

    Video individual counting for moving drones,

    Y . Fan, J. Wan, T. Han, A. B. Chan, and A. J. Ma, “Video individual counting for moving drones,” inProceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV), October 2025, pp. 12 284–12 293

    2025

  29. [37]

    Density-based flow mask in- tegration via deformable convolution for video people flux estimation,

    C.-L. Wan, F.-K. Huang, and H.-H. Shuai, “Density-based flow mask in- tegration via deformable convolution for video people flux estimation,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), January 2024, pp. 6573–6582

    2024

  30. [38]

    Flowing crowd to count flows: A self-supervised frame- work for video individual counting,

    F.-K. Huang, B.-L. Huang, L.-W. Tsao, J.-C. Wu, H.-H. Shuai, and W.-H. Cheng, “Flowing crowd to count flows: A self-supervised frame- work for video individual counting,” ser. MM ’25. Association for Computing Machinery, 2025, p. 8234–8243

    2025

  31. [39]

    Context-aware crowd counting,

    W. Liu, M. Salzmann, and P. Fua, “Context-aware crowd counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2019

    2019

  32. [40]

    Reverse perspec- tive network for perspective-aware object counting,

    Y . Yang, G. Li, Z. Wu, L. Su, Q. Huang, and N. Sebe, “Reverse perspec- tive network for perspective-aware object counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2020

    2020

  33. [41]

    Distribution matching for crowd counting,

    B. Wang, H. Liu, D. Samaras, and M. Hoai, “Distribution matching for crowd counting,” inAdvances in Neural Information Processing Systems, 2020

    2020

  34. [42]

    Steerer: Resolving scale varia- tions for counting and localization via selective inheritance learning,

    T. Han, L. Bai, L. Liu, and W. Ouyang, “Steerer: Resolving scale varia- tions for counting and localization via selective inheritance learning,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2023, pp. 21 848–21 859

    2023

  35. [43]

    Crowddiff: Multi-hypothesis crowd density estimation using diffusion models,

    Y . Ranasinghe, N. G. Nair, W. G. C. Bandara, and V . M. Patel, “Crowddiff: Multi-hypothesis crowd density estimation using diffusion models,” inCVPR, 2024, pp. 12 809–12 819

    2024

  36. [44]

    Face recognition using kernel ridge regression,

    S. An, W. Liu, and S. Venkatesh, “Face recognition using kernel ridge regression,” in2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007, pp. 1–7

    2007

  37. [45]

    Crowd density estimation using texture analysis and learning,

    X. Wu, G. Liang, K. K. Lee, and Y . Xu, “Crowd density estimation using texture analysis and learning,” in2006 IEEE International Conference on Robotics and Biomimetics, 2006, pp. 214–219

    2006

  38. [46]

    Privacy preserving crowd monitoring: Counting people without people models or tracking,

    A. B. Chan, Z.-S. J. Liang, and N. Vasconcelos, “Privacy preserving crowd monitoring: Counting people without people models or tracking,” in2008 IEEE Conference on Computer Vision and Pattern Recognition, 2008, pp. 1–7

    2008

  39. [47]

    Deep people counting in extremely dense crowds,

    C. Wang, H. Zhang, L. Yang, S. Liu, and X. Cao, “Deep people counting in extremely dense crowds,” inProceedings of the 23rd ACM International Conference on Multimedia, ser. MM ’15. Association for Computing Machinery, 2015, p. 1299–1302

    2015

  40. [48]

    Revisiting perspective infor- mation for efficient crowd counting,

    M. Shi, Z. Yang, C. Xu, and Q. Chen, “Revisiting perspective infor- mation for efficient crowd counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2019

    2019

  41. [49]

    Perspective-guided convolution networks for crowd counting,

    Z. Yan, Y . Yuan, W. Zuo, X. Tan, Y . Wang, S. Wen, and E. Ding, “Perspective-guided convolution networks for crowd counting,” inPro- ceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2019

    2019

  42. [50]

    Redesigning multi-scale neural network for crowd counting,

    Z. Du, M. Shi, J. Deng, and S. Zafeiriou, “Redesigning multi-scale neural network for crowd counting,”IEEE Transactions on Image Processing, vol. 32, pp. 3664–3678, 2023

    2023

  43. [51]

    Stnet: Scale tree network with multi-level auxiliator for crowd counting,

    M. Wang, H. Cai, X.-F. Han, J. Zhou, and M. Gong, “Stnet: Scale tree network with multi-level auxiliator for crowd counting,”IEEE Transactions on Multimedia, vol. 25, pp. 2074–2084, 2023

  44. [52]

    Domain-adaptive crowd counting via high-quality image translation and density reconstruction,

    J. Gao, T. Han, Y . Yuan, and Q. Wang, “Domain-adaptive crowd counting via high-quality image translation and density reconstruction,” IEEE Transactions on Neural Networks and Learning Systems, vol. 34, no. 8, pp. 4803–4815, 2023

    2023

  45. [53]

    Domain-general crowd counting in unseen scenarios,

    Z. Du, J. Deng, and M. Shi, “Domain-general crowd counting in unseen scenarios,”Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 1, pp. 561–570, Jun. 2023

    2023

  46. [54]

    Striking a balance: Unsu- pervised cross-domain crowd counting via knowledge diffusion,

    H. Xie, Z. Yang, H. Zhu, and Z. Wang, “Striking a balance: Unsu- pervised cross-domain crowd counting via knowledge diffusion,” in Proceedings of the 31st ACM International Conference on Multimedia, 2023, p. 6520–6529

    2023

  47. [55]

    Bi-level alignment for cross-domain crowd counting,

    S. Gong, S. Zhang, J. Yang, D. Dai, and B. Schiele, “Bi-level alignment for cross-domain crowd counting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 7542–7550

    2022

  48. [56]

    A generalized loss function for crowd counting and localization,

    J. Wan, Z. Liu, and A. B. Chan, “A generalized loss function for crowd counting and localization,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 1974–1983

    2021

  49. [57]

    Bayesian poisson regression for crowd counting,

    A. B. Chan and N. Vasconcelos, “Bayesian poisson regression for crowd counting,” in2009 IEEE 12th International Conference on Computer Vision, 2009, pp. 545–551

    2009

  50. [58]

    An end-to-end transformer model for crowd localization,

    D. Liang, W. Xu, and X. Bai, “An end-to-end transformer model for crowd localization,” inProceedings of the European Conference on Computer Vision (ECCV), 2022, pp. 38–54

    2022

  51. [59]

    Learning crowd scale and distribution for weakly supervised crowd counting and localization,

    Y . Fan, J. Wan, and A. J. Ma, “Learning crowd scale and distribution for weakly supervised crowd counting and localization,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 35, no. 1, pp. 713– 727, 2025

    2025

  52. [60]

    Delving into the trajectory long-tail distribution for muti-object tracking,

    S. Chen, E. Yu, J. Li, and W. Tao, “Delving into the trajectory long-tail distribution for muti-object tracking,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 19 341–19 351

    2024

  53. [61]

    Deconfusetrack: Dealing with confusion for multi-object tracking,

    C. Huang, S. Han, M. He, W. Zheng, and Y . Wei, “Deconfusetrack: Dealing with confusion for multi-object tracking,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 19 290–19 299

    2024

  54. [62]

    Utm: A unified multiple object tracking model with identity-aware feature enhancement,

    S. You, H. Yao, B.-K. Bao, and C. Xu, “Utm: A unified multiple object tracking model with identity-aware feature enhancement,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2023, pp. 21 876–21 886

    2023

  55. [63]

    Bot-sort: Robust asso- ciations multi-pedestrian tracking,

    N. Aharon, R. Orfaig, and B.-Z. Bobrovsky, “Bot-sort: Robust asso- ciations multi-pedestrian tracking,”arXiv preprint arXiv:2206.14651, 2022

    2022 arXiv

  56. [64]

    Diffmot: A real-time diffusion-based multiple object tracker with non-linear prediction,

    W. Lv, Y . Huang, N. Zhang, R.-S. Lin, M. Han, and D. Zeng, “Diffmot: A real-time diffusion-based multiple object tracker with non-linear prediction,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 19 321–19 330

    2024

  57. [65]

    Observation- centric sort: Rethinking sort for robust multi-object tracking,

    J. Cao, J. Pang, X. Weng, R. Khirodkar, and K. Kitani, “Observation- centric sort: Rethinking sort for robust multi-object tracking,” inPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 9686–9696

    2023

  58. [66]

    A new approach to linear filtering and prediction problems,

    R. E. Kalman, “A new approach to linear filtering and prediction problems,”Journal of Basic Engineering, vol. 82, no. 1, pp. 35–45, 03 1960

    1960

  59. [67]

    Motr: End-to-end multiple-object tracking with transformer,

    F. Zeng, B. Dong, Y . Zhang, T. Wang, X. Zhang, and Y . Wei, “Motr: End-to-end multiple-object tracking with transformer,” inComputer Vision – ECCV 2022, 2022, pp. 659–675

    2022

  60. [68]

    Rgb-t crowd counting from drone: A benchmark and mmccn network,

    T. Peng, Q. Li, and P. Zhu, “Rgb-t crowd counting from drone: A benchmark and mmccn network,” inComputer Vision – ACCV 2020, 2021, pp. 497–513

    2020

  61. [69]

    Drone- based joint density map estimation, localization and tracking with space- time multi-scale attention network,

    L. Wen, D. Du, P. Zhu, Q. Hu, Q. Wang, L. Bo, and S. Lyu, “Drone- based joint density map estimation, localization and tracking with space- time multi-scale attention network,”arXiv preprint arXiv:1912.01811, 2019

    1912 arXiv

  62. [70]

    A large- scale drone based thermal infrared benchmark and inception transformer network for crowd counting,

    X. Wang, T. Li, Y . Liu, S. Yao, Y . Liu, N. Yang, and P. Zhu, “A large- scale drone based thermal infrared benchmark and inception transformer network for crowd counting,”Pattern Recognition, vol. 173, p. 112778, 2026

    2026

  63. [71]

    Enhanced uav-dot for uav crowd localization: Adaptive gaussian heat map and attention mechanism to address scale/low-light challenges,

    M. Zhang, F. Zhao, and Y . Zhang, “Enhanced uav-dot for uav crowd localization: Adaptive gaussian heat map and attention mechanism to address scale/low-light challenges,”Drones, vol. 9, no. 12, 2025

    2025

  64. [72]

    Drone-person tracking in uniform appearance crowd: A new dataset,

    M. Alansari, O. A. Hay, S. Alansari, S. Javed, A. Shoufan, Y . Zweiri, and N. Werghi, “Drone-person tracking in uniform appearance crowd: A new dataset,”Scientific Data, vol. 11, no. 1, p. 15, 2024

    2024

  65. [73]

    Dense- track: Drone-based crowd tracking via density-aware motion-appearance IEEE TRANSACTIONS 16 synergy,

    Y . Lei, H. Zhu, J. Yuan, G. Xiang, X. Zhong, and S. He, “Dense- track: Drone-based crowd tracking via density-aware motion-appearance IEEE TRANSACTIONS 16 synergy,” inProceedings of the 32nd ACM International Conference on Multimedia. Association for Computing Machinery, 2024...

    2024

  66. [74]

    Multi-frame attention with feature-level warping for drone crowd tracking,

    T. Asanomi, K. Nishimura, and R. Bise, “Multi-frame attention with feature-level warping for drone crowd tracking,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023, pp. 1664–1673

    2023

  67. [75]

    Tracking pedestrian heads in dense crowd,

    R. Sundararaman, C. De Almeida Braga, E. Marchand, and J. Pettre, “Tracking pedestrian heads in dense crowd,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2021, pp. 3865–3875

    2021

  68. [76]

    Superglue: Learning feature matching with graph neural networks,

    P.-E. Sarlin, D. DeTone, T. Malisiewicz, and A. Rabinovich, “Superglue: Learning feature matching with graph neural networks,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition (CVPR), June 2020

    2020

  69. [77]

    Sinkhorn distances: Lightspeed computation of optimal transport,

    M. Cuturi, “Sinkhorn distances: Lightspeed computation of optimal transport,” inAdvances in Neural Information Processing Systems, vol. 26, 2013

    2013

  70. [78]

    Algorithms for the assignment and transportation prob- lems,

    J. Munkres, “Algorithms for the assignment and transportation prob- lems,”Journal of the Society for Industrial and Applied Mathematics, vol. 5, no. 1, pp. 32–38, 1957

    1957

  71. [79]

    Hota: A higher order metric for evaluating multi-object tracking,

    J. Luiten, A. O ˇsep, P. Dendorfer, P. Torr, A. Geiger, L. Leal-Taix ´e, and B. Leibe, “Hota: A higher order metric for evaluating multi-object tracking,”International Journal of Computer Vision, vol. 129, no. 2, pp. 548–578, 2021

    2021

  72. [80]

    Evaluating multiple object tracking performance: the clear mot metrics,

    K. Bernardin and R. Stiefelhagen, “Evaluating multiple object tracking performance: the clear mot metrics,”EURASIP Journal on Image and Video Processing, vol. 2008, no. 1, p. 246309, 2008

    2008

  73. [81]

    Perfor- mance measures and a data set for multi-target, multi-camera tracking,

    E. Ristani, F. Solera, R. Zou, R. Cucchiara, and C. Tomasi, “Perfor- mance measures and a data set for multi-target, multi-camera tracking,” inComputer Vision – ECCV 2016 Workshops, G. Hua and H. J ´egou, Eds. Cham: Springer International Publishing, 2016, pp. 17–35

    2016

  74. [82]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016

    2016

  75. [83]

    Feature pyramid networks for object detection,

    T.-Y . Lin, P. Dollar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), July 2017

    2017

  76. [84]

    Crossing-line crowd counting with two-phase deep neural networks,

    Z. Zhao, H. Li, R. Zhao, and X. Wang, “Crossing-line crowd counting with two-phase deep neural networks,” inComputer Vision – ECCV 2016, 2016, pp. 712–726

    2016

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.