Pith. sign in

REVIEW 3 major objections 3 minor 69 references

SocialTrack: Multi-Object Tracking in Complex Urban Traffic Scenes Inspired by Social Behavior

T0 review · 3 major / 3 minor · reviewed 2026-08-05 · deepseek-v4-flash

Pith's one-line read SocialTrack claims that adding velocity-adaptive filtering, social-group motion references, and trajectory memory to a multi-object tracker lifts tracking accuracy and identity consistency above existing state-of-the-art on drone traffic fo

desk verdict The submitted full text is for a different paper (HeteroRAG), so SocialTrack's claimed MOTA/IDF1 gains are completely unsupported by the provided manuscript. read the letter →

arxiv 2508.12777 v3 pith:RHZKJFFR submitted 2025-08-18 cs.CV

classification cs.CV
keywords multi-objecttrackingUAVimageryurbantrafficsmall-targetdetectionKalmanfiltergroupmotioncompensationspatio-temporalmemoryidentityswitching
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

SocialTrack is a modular multi-object tracking framework aimed at small, low-quality targets in complex urban traffic filmed from drones. It claims that four additions—a small-target detector with multi-scale feature enhancement, a velocity-adaptive Kalman filter (VACKF), a group-motion compensation strategy (GMCS), and a spatio-temporal memory predictor (STMP)—combine to raise tracking accuracy and reduce identity switches. On the UAVDT and MOT17 benchmarks, the framework is reported to outperform existing state-of-the-art trackers on the standard MOTA and IDF1 metrics (tracking accuracy and identity consistency). If true, the practical payoff is more stable tracking of small vehicles through occlusion, crossing, and blur in drone-based traffic monitoring.

What carries the argument

The framework is carried by four named mechanisms: (1) a multi-scale feature enhancement small-target detector, which improves detection of tiny objects; (2) the Velocity Adaptive Cubature Kalman Filter (VACKF), a trajectory predictor whose motion model adapts using measured velocity; (3) the Group Motion Compensation Strategy (GMCS), which encodes social group motion priors to give low-quality tracks stable state-update references; and (4) the Spatio-Temporal Memory Prediction (STMP), which forecasts the future state of low-quality tracks from historical trajectories. The paper positions GMCS as the module responsible for improved association accuracy and STMP as the module responsible for

What would settle it

Run SocialTrack on UAVDT and MOT17 with GMCS disabled (low-quality tracks get no social-motion reference) while keeping the other three modules intact. If MOTA, IDF1, and identity-switch counts stay essentially unchanged, the social prior is not carrying the claimed association gain; a second probe is to apply the full framework to an unseen drone-traffic dataset and compare identity-switch rates against the reported values.

Watch

Extended reading notes

Core claim

The central claim is that the failure modes of UAV multi-object tracking—small-scale variations, occlusion, nonlinear crossing motion, and motion blur—can be attacked at the motion-model and association level, not just by a stronger detector. The paper argues that a dedicated small-target detector supplies better detections, while VACKF replaces fixed filtering with velocity-adaptive dynamics; GMCS uses the prior that nearby targets move as a social group to stabilize state updates for low-quality tracks; and STMP uses historical trajectories to predict where such tracks will reappear, cutting identity switches. Together, these components are claimed to yield significant gains in MOTA and ID

Load-bearing premise

The load-bearing premise is that the social group-motion prior (GMCS) matches how targets actually move in the UAVDT and MOT17 benchmarks—and that those two datasets represent the claimed urban-traffic setting—so low-quality tracks receive state updates that improve association instead of adding bias.

Editorial extensions

If this is right

  • If the claims hold, drone-based traffic analysts can track small vehicles longer through occlusions and crossing maneuvers with fewer identity switches.
  • The modular design implies existing trackers can gain accuracy by swapping in VACKF, GMCS, or STMP without retraining a complete system.
  • Velocity-adaptive filtering should improve trajectory continuity on nonlinear and fast-moving targets, not only on small ones.
  • Reported gains on both UAVDT and the non-UAV MOT17 benchmark would indicate that the modules transfer beyond drone footage to ordinary street scenes.

Reading between the lines

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

  • The supplied full text for this record is a different manuscript (a medical retrieval paper), so the component-level ablations that would demonstrate each module's contribution are not available here; the summary relies on the SocialTrack abstract.
  • If the social-group prior is the active ingredient, applying GMCS only after ego-motion compensation—or only within localized traffic corridors—should outperform applying it globally, a configuration the paper leaves implicit.
  • A natural extension is to combine STMP with appearance-based re-identification, letting the memory predictor propose candidate states while appearance matching confirms them; this could push IDF1 further than either mechanism alone.
  • Running the four modules on an unseen drone-traffic benchmark would test whether the reported gains generalize beyond the two datasets used in the paper.
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

3 major / 3 minor

Summary. The submission, arXiv:2508.12777 (SocialTrack), presents an abstract claiming a multi-object tracking framework for complex urban UAV traffic scenes. The framework is said to combine a specialized small-target detector, a Velocity Adaptive Cubature Kalman Filter (VACKF), a Group Motion Compensation Strategy (GMCS), and a Spatio-Temporal Memory Prediction (STMP) module, and is claimed to outperform state-of-the-art methods on UAVDT and MOT17 in MOTA and IDF1. However, the full text supplied with the submission is not SocialTrack but HeteroRAG, a medical retrieval-augmented generation paper on an unrelated topic. Consequently, the submitted manuscript contains no description, equations, experiments, ablations, or results for SocialTrack; the only evidence for the central claim is the abstract itself.

Significance. If the claimed results were supported, the work could be of practical interest to UAV-based multi-object tracking and could contribute modular components for existing trackers. However, the supplied manuscript provides no verifiable technical content for SocialTrack. There are no definitions of the four proposed components, no derivations of the filter or motion-prior models, no benchmark protocol, no result tables, and no code or reproducibility artifacts. The claimed SOTA improvements on MOTA and IDF1 are therefore not assessable. The paper cannot be credited with a verifiable contribution in its current form.

major comments (3)
  1. [Full text (all sections)] The full text is a different paper: it is HeteroRAG, a medical vision-language retrieval-augmented generation paper (arXiv:2508.12778), and contains none of the SocialTrack content. The proposed small-target detector, VACKF, GMCS, and STMP are never described; there are no equations, no UAVDT/MOT17 experiments, and no ablation tables. The central claim that SocialTrack outperforms SOTA on MOTA and IDF1 rests entirely on the abstract's assertion. This is a load-bearing failure of evidence, not a local presentation issue.
  2. [Abstract] Each of the four claimed components is introduced only by name. No mathematical formulation of the velocity-adaptive cubature Kalman filter, the group-motion compensation prior, or the spatio-temporal memory prediction is supplied, so the reader cannot check whether the motion priors are well-posed, whether the filter is derived consistently, or whether the proposed mechanisms would reduce bias in low-quality track state updates. These definitions are prerequisites for evaluating the paper's central technical claims.
  3. [Abstract (evaluation)] The claimed evaluation on UAVDT and MOT17 is unverifiable: no protocol, metric definitions, baseline list, result table, or error bars appear in the supplied manuscript. Moreover, MOT17 is primarily a static-camera benchmark, which conflicts with the stated focus on complex UAV perspectives; whether this mismatch undermines the claimed generalization cannot be assessed because the experimental section is absent.
minor comments (3)
  1. [Metadata and full text] The submission metadata identifies the paper as SocialTrack (cs.CV, 2508.12777), but the full text is HeteroRAG (cs.CL, 2508.12778) with different authors and content. This mismatch should be resolved by the authors before any further consideration.
  2. [Abstract] The abstract gives no references to prior SOTA methods or to the UAVDT/MOT17 benchmarks, making it impossible to position the claimed improvements in context.
  3. [Reproducibility] No code, model weights, or detailed reproducibility statement is provided for SocialTrack; the only such artifacts in the supplied text belong to HeteroRAG.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity detectable: the SocialTrack abstract contains no derivation chain, and the supplied full text is a different paper; unsupported claims are an evidence problem, not circularity.

full rationale

The target manuscript is only represented by the SocialTrack abstract; the supplied full text is entirely the HeteroRAG medical retrieval-augmented generation paper (arXiv:2508.12778), not SocialTrack. The SocialTrack abstract makes claims about a small-target detector, VACKF, GMCS, and STMP improving MOTA/IDF1 on UAVDT and MOT17, but it contains no equations, no fitted parameters, no derivations, and no citation chain that could reduce a 'prediction' to its inputs. There is therefore no load-bearing circular step to exhibit: no Eq. X = Eq. Y by construction, no fitted parameter renamed as a prediction, and no self-citation invoked as a substitute for derivation. The mismatch between abstract and body is a serious evidence-integrity problem, because the experiments and ablations that would substantiate the claimed SOTA results are absent from the text supplied for review. However, absence of evidence is not circularity, and the instructions prohibit claiming circularity without quoting a specific reduction. Accordingly, the appropriate circularity score is 0, while the supportability of the central claim remains unassessable and should be treated as a separate correctness/evidence concern.

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

The abstract alone does not expose explicit free parameters or new physical entities. The central modeling assumptions concern the validity of social group motion priors, the representativeness of the benchmarks, and the framing of the listed challenges as the primary bottlenecks. Full verification requires the missing SocialTrack manuscript.

assumptions (3)
  • domain assumption The challenges listed (small target scale variations, occlusions, nonlinear crossing motions, motion blur) are the dominant factors limiting MOT performance in complex UAV scenes.
    Stated in the abstract as the motivation for the framework; the paper does not provide evidence in the abstract that these are the main bottlenecks relative to other factors.
  • domain assumption A social group motion prior (GMCS) provides a stable and useful state update reference for low-quality tracks in urban traffic.
    This is the core modeling assumption of the GMCS component; its validity is asserted rather than justified in the abstract.
  • domain assumption The UAVDT and MOT17 benchmarks are representative of complex urban traffic scenes and their metrics capture the claimed robustness improvements.
    The abstract reports SOTA on these datasets; MOT17 is not a UAV dataset, so the transferability claim rests on an assumption about benchmark representativeness.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SocialTrack: Multi-Object Tracking in Complex Urban Traffic Scenes Inspired by Social Behavior." pith.science (2026). https://pith.science/paper/RHZKJFFR

@misc{pith2026250812777,
  author       = {Pith},
  title        = {Pith review of: SocialTrack: Multi-Object Tracking in Complex Urban Traffic Scenes Inspired by Social Behavior},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RHZKJFFR}},
  note         = {Machine review of arXiv:2508.12777}
}
read the original abstract

As a key research direction in the field of multi-object tracking (MOT), UAV-based multi-object tracking has significant application value in the analysis and understanding of urban intelligent transportation systems. However, in complex UAV perspectives, challenges such as small target scale variations, occlusions, nonlinear crossing motions, and motion blur severely hinder the stability of multi-object tracking. To address these challenges, this paper proposes a novel multi-object tracking framework, SocialTrack, aimed at enhancing the tracking accuracy and robustness of small targets in complex urban traffic environments. The specialized small-target detector enhances the detection performance by employing a multi-scale feature enhancement mechanism. The Velocity Adaptive Cubature Kalman Filter (VACKF) improves the accuracy of trajectory prediction by incorporating a velocity dynamic modeling mechanism. The Group Motion Compensation Strategy (GMCS) models social group motion priors to provide stable state update references for low-quality tracks, significantly improving the target association accuracy in complex dynamic environments. Furthermore, the Spatio-Temporal Memory Prediction (STMP) leverages historical trajectory information to predict the future state of low-quality tracks, effectively mitigating identity switching issues. Extensive experiments on the UAVDT and MOT17 datasets demonstrate that SocialTrack outperforms existing state-of-the-art (SOTA) methods across several key metrics. Significant improvements in MOTA and IDF1, among other core performance indicators, highlight its superior robustness and adaptability. Additionally, SocialTrack is highly modular and compatible, allowing for seamless integration with existing trackers to further enhance performance.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

69 extracted references · 66 canonical work pages

  1. [1]

    arXiv preprint arXiv:220614651

    Aharon N, Orfaig R, Bobrovsky BZ (2022) Bot-sort: Robust associations multi-pedestrian tracking. arXiv preprint arXiv:220614651

  2. [2]

    IEEE Transactions on automatic control 54(6):1254--1269

    Arasaratnam I, Haykin S (2009) Cubature kalman filters. IEEE Transactions on automatic control 54(6):1254--1269

  3. [3]

    Computers and Electronics in Agriculture 219:108757

    Ariza-Sent \' s M, V \'e lez S, Mart \' nez-Pe \ n a R, et al (2024) Object detection and tracking in precision farming: A systematic review. Computers and Electronics in Agriculture 219:108757

  4. [4]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1218--1225

    Bae SH, Yoon KJ (2014) Robust online multi-object tracking based on tracklet confidence and online discriminative appearance learning. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1218--1225

  5. [5]

    EURASIP Journal on Image and Video Processing 2008:1--10

    Bernardin K, Stiefelhagen R (2008) Evaluating multiple object tracking performance: the clear mot metrics. EURASIP Journal on Image and Video Processing 2008:1--10

  6. [6]

    In: 2016 IEEE international conference on image processing (ICIP), Ieee, pp 3464--3468

    Bewley A, Ge Z, Ott L, et al (2016) Simple online and realtime tracking. In: 2016 IEEE international conference on image processing (ICIP), Ieee, pp 3464--3468

  7. [7]

    Artificial Intelligence Review 58(10):309

    Bi J, Dornaika F, Charafeddine J (2025) Linear projection fused graph-based semi-supervised learning on multi-view data. Artificial Intelligence Review 58(10):309

  8. [8]

    In: 2017 14th IEEE international conference on advanced video and signal based surveillance (AVSS), IEEE, pp 1--6

    Bochinski E, Eiselein V, Sikora T (2017) High-speed tracking-by-detection without using image information. In: 2017 14th IEEE international conference on advanced video and signal based surveillance (AVSS), IEEE, pp 1--6

Show all 69 references
  1. [9]

    In: 2018 IEEE international conference on robotics and automation (ICRA), IEEE, pp 2064--2069

    Buyval A, Gabdullin A, Mustafin R, et al (2018) Realtime vehicle and pedestrian tracking for didi udacity self-driving car challenge. In: 2018 IEEE international conference on robotics and automation (ICRA), IEEE, pp 2064--2069

  2. [10]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 8090--8100

    Cai J, Xu M, Li W, et al (2022) Memot: Multi-object tracking with memory. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 8090--8100

  3. [11]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 9686--9696

    Cao J, Pang J, Weng X, et al (2023) Observation-centric sort: Rethinking sort for robust multi-object tracking. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 9686--9696

  4. [12]

    In: European conference on computer vision, Springer, pp 213--229

    Carion N, Massa F, Synnaeve G, et al (2020) End-to-end object detection with transformers. In: European conference on computer vision, Springer, pp 213--229

  5. [13]

    In: 2018 IEEE international conference on multimedia and expo (ICME), IEEE, pp 1--6

    Chen L, Ai H, Zhuang Z, et al (2018) Real-time multiple people tracking with deeply learned candidate selection and person re-identification. In: 2018 IEEE international conference on multimedia and expo (ICME), IEEE, pp 1--6

  6. [14]

    In: Proceedings of the IEEE international conference on computer vision, pp 2304--2311

    Dicle C, Camps OI, Sznaier M (2013) The way they move: Tracking multiple targets with similar appearance. In: Proceedings of the IEEE international conference on computer vision, pp 2304--2311

  7. [15]

    In: Proceedings of the European conference on computer vision (ECCV), pp 370--386

    Du D, Qi Y, Yu H, et al (2018) The unmanned aerial vehicle benchmark: Object detection and tracking. In: Proceedings of the European conference on computer vision (ECCV), pp 370--386

  8. [16]

    ACM Computing Surveys (CSUR) 53(4):1--34

    Emami P, Pardalos PM, Elefteriadou L, et al (2020) Machine learning methods for data association in multi-object tracking. ACM Computing Surveys (CSUR) 53(4):1--34

  9. [17]

    Sensors 23(8):3852

    Fei L, Han B (2023) Multi-object multi-camera tracking based on deep learning for intelligent transportation: A review. Sensors 23(8):3852

  10. [18]

    Artificial Intelligence Review 56(Suppl 1):1417--1477

    Fu C, Lu K, Zheng G, et al (2023) Siamese object tracking for unmanned aerial vehicle: A review and comprehensive analysis. Artificial Intelligence Review 56(Suppl 1):1417--1477

  11. [19]

    arXiv preprint arXiv:210708430

    Ge Z, Liu S, Wang F, et al (2021) Yolox: Exceeding yolo series in 2021. arXiv preprint arXiv:210708430

  12. [20]

    Artificial Intelligence Review 58(8):235

    Guan Z, Wang Z, Zhang G, et al (2025) Multi-object tracking review: retrospective and emerging trend. Artificial Intelligence Review 58(8):235

  13. [21]

    Advances in neural information processing systems 35:8291--8303

    Han K, Wang Y, Guo J, et al (2022) Vision gnn: An image is worth graph of nodes. Advances in neural information processing systems 35:8291--8303

  14. [22]

    Multimedia Tools and Applications 83(14):43439--43492

    Hassan S, Mujtaba G, Rajput A, et al (2024) Multi-object tracking: a systematic literature review. Multimedia Tools and Applications 83(14):43439--43492

  15. [23]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 770--778

    He K, Zhang X, Ren S, et al (2016) Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 770--778

  16. [24]

    Artificial Intelligence Review 58(6):1--67

    Hua W, Chen Q (2025) A survey of small object detection based on deep learning in aerial images. Artificial Intelligence Review 58(6):1--67

  17. [25]

    arXiv preprint arXiv:240310830

    Ji D, Gao S, Zhu L, et al (2024) View-centric multi-object tracking with homographic matching in moving uav. arXiv preprint arXiv:240310830

  18. [26]

    Artificial Intelligence Review 56(Suppl 3):3515--3570

    Jiao R, Wan Y, Poiesi F, et al (2023) Survey on video anomaly detection in dynamic scenes with moving cameras. Artificial Intelligence Review 56(Suppl 3):3515--3570

  19. [27]

    Advanced Engineering Informatics 59:102250

    Jiao Y, Zhai X, Peng L, et al (2024) A digital twin-based motion forecasting framework for preemptive risk monitoring. Advanced Engineering Informatics 59:102250

  20. [28]

    Naval research logistics quarterly 2(1-2):83--97

    Kuhn HW (1955) The hungarian method for the assignment problem. Naval research logistics quarterly 2(1-2):83--97

  21. [29]

    IEEE Transactions on Intelligent Transportation Systems

    Li H, Liu H, Du Z, et al (2024) Mcca-mot: Multimodal collaboration-guided cascade association network for 3d multi-object tracking. IEEE Transactions on Intelligent Transportation Systems

  22. [30]

    Drones 7(11):681

    Li X, Zhu R, Yu X, et al (2023) High-performance detection-based tracker for multiple object tracking in uavs. Drones 7(11):681

  23. [31]

    IEEE Transactions on Circuits and Systems for Video Technology

    Lin J, Liang G, Zhang R (2024) Lttrack: Rethinking the tracking framework for long-term multi-object tracking. IEEE Transactions on Circuits and Systems for Video Technology

  24. [32]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 9996--10005

    Liu K, Jin S, Fu Z, et al (2023 a ) Uncertainty-aware unsupervised multi-object tracking. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 9996--10005

  25. [33]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 6027--6037

    Liu W, Lu H, Fu H, et al (2023 b ) Learning to upsample by learning to sample. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 6027--6037

  26. [34]

    IEEE Transactions on Intelligent Transportation Systems

    Liu Y, Liu X, Jiang Z, et al (2025 a ) Co-mot: Exploring the collaborative relations in traffic flow for 3d multi-object tracking. IEEE Transactions on Intelligent Transportation Systems

  27. [35]

    IEEE Transactions on Multimedia 25:1462--1476

    Liu Z, Shang Y, Li T, et al (2023 c ) Robust multi-drone multi-target tracking to resolve target occlusion: A benchmark. IEEE Transactions on Multimedia 25:1462--1476

  28. [36]

    IEEE Transactions on Circuits and Systems for Video Technology

    Liu Z, Wang X, Wang C, et al (2025 b ) Sparsetrack: Multi-object tracking by performing scene decomposition based on pseudo-depth. IEEE Transactions on Circuits and Systems for Video Technology

  29. [37]

    International journal of computer vision 129:548--578

    Luiten J, Osep A, Dendorfer P, et al (2021) Hota: A higher order metric for evaluating multi-object tracking. International journal of computer vision 129:548--578

  30. [38]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 8844--8854

    Meinhardt T, Kirillov A, Leal-Taixe L, et al (2022) Trackformer: Multi-object tracking with transformers. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 8844--8854

  31. [39]

    arXiv preprint arXiv:160300831

    Milan A, Leal-Taix \'e L, Reid I, et al (2016) Mot16: A benchmark for multi-object tracking. arXiv preprint arXiv:160300831

  32. [40]

    In: 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp 5224--5231

    MohaimenianPour S, Vaughan R (2018) Hands and faces, fast: mono-camera user detection robust enough to directly control a uav in flight. In: 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp 5224--5231

  33. [41]

    Drones 6(6):147

    Mohsan SAH, Khan MA, Noor F, et al (2022) Towards the unmanned aerial vehicles (uavs): A comprehensive review. Drones 6(6):147

  34. [42]

    Artificial Intelligence Review 58(7):202

    Nasir R, Jalil Z, Nasir M, et al (2025) An enhanced framework for real-time dense crowd abnormal behavior detection using yolov8. Artificial Intelligence Review 58(7):202

  35. [43]

    IEEE Transactions on Multimedia 26:972--983

    Nguyen TT, Nguyen HH, Sartipi M, et al (2023) Multi-vehicle multi-camera tracking with graph-based tracklet features. IEEE Transactions on Multimedia 26:972--983

  36. [44]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 164--173

    Pang J, Qiu L, Li X, et al (2021) Quasi-dense similarity learning for multiple object tracking. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 164--173

  37. [45]

    In: Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part IV 16, Springer, pp 145--161

    Peng J, Wang C, Wan F, et al (2020) Chained-tracker: Chaining paired attentive regression results for end-to-end joint multiple-object detection and tracking. In: Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part IV 16, S...

  38. [46]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 11289--11298

    Ren H, Han S, Ding H, et al (2023) Focus on details: Online multi-object tracking with diverse fine-grained representation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 11289--11298

  39. [47]

    IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

    Ren L, Yin W, Diao W, et al (2025) Supermot: Decoupling motion and fusing temporal pyramid features for uav multi-object tracking. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

  40. [48]

    Ronneberger O, Fischer P, Brox T (2015) U-net: Convolutional networks for biomedical image segmentation. In: Medical image computing and computer-assisted intervention--MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18, Sp...

  41. [49]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 13813--13823

    Seidenschwarz J, Bras \'o G, Serrano VC, et al (2023) Simple cues lead to a strong multi-object tracker. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 13813--13823

  42. [50]

    In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp 10870--10877

    Song I, Lee J (2024) Sftrack: A robust scale and motion adaptive algorithm for tracking small and fast moving objects. In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp 10870--10877

  43. [51]

    Engineering Applications of Artificial Intelligence 119:105770

    Tsai CY, Shen GY, Nisar H (2023) Swin-jde: Joint detection and embedding multi-object tracking in crowded scenes based on swin-transformer. Engineering Applications of Artificial Intelligence 119:105770

  44. [52]

    IEEE transactions on pattern analysis and machine intelligence 44(7):3614--3633

    Vandenhende S, Georgoulis S, Van Gansbeke W, et al (2021) Multi-task learning for dense prediction tasks: A survey. IEEE transactions on pattern analysis and machine intelligence 44(7):3614--3633

  45. [53]

    IEEE Transactions on Intelligent Transportation Systems 24(11):11981--11996

    Wang L, Zhang X, Qin W, et al (2023) Camo-mot: Combined appearance-motion optimization for 3d multi-object tracking with camera-lidar fusion. IEEE Transactions on Intelligent Transportation Systems 24(11):11981--11996

  46. [54]

    IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

    Wang Q, Zhou L, Jin P, et al (2024) Trackingmamba: Visual state space model for object tracking. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

  47. [55]

    In: 2017 IEEE international conference on image processing (ICIP), IEEE, pp 3645--3649

    Wojke N, Bewley A, Paulus D (2017) Simple online and realtime tracking with a deep association metric. In: 2017 IEEE international conference on image processing (ICIP), IEEE, pp 3645--3649

  48. [56]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 12352--12361

    Wu J, Cao J, Song L, et al (2021) Track to detect and segment: An online multi-object tracker. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 12352--12361

  49. [57]

    In: Proceedings of the IEEE international conference on computer vision, pp 4705--4713

    Xiang Y, Alahi A, Savarese S (2015) Learning to track: Online multi-object tracking by decision making. In: Proceedings of the IEEE international conference on computer vision, pp 4705--4713

  50. [58]

    In: Proceedings of the IEEE/CVF winter conference on applications of computer vision, pp 4799--4808

    Yang F, Odashima S, Masui S, et al (2023) Hard to track objects with irregular motions and similar appearances? make it easier by buffering the matching space. In: Proceedings of the IEEE/CVF winter conference on applications of computer vision, pp 4799--4808

  51. [59]

    In: Proceedings of the AAAI conference on artificial intelligence, pp 6702--6710

    Yi K, Luo K, Luo X, et al (2024) Ucmctrack: Multi-object tracking with uniform camera motion compensation. In: Proceedings of the AAAI conference on artificial intelligence, pp 6702--6710

  52. [60]

    Yu W, Wang X (2025) Mambaout: Do we really need mamba for vision? In: Proceedings of the Computer Vision and Pattern Recognition Conference, pp 4484--4496

  53. [61]

    Neural computation 31(7):1235--1270

    Yu Y, Si X, Hu C, et al (2019) A review of recurrent neural networks: Lstm cells and network architectures. Neural computation 31(7):1235--1270

  54. [62]

    In: European conference on computer vision, Springer, pp 659--675

    Zeng F, Dong B, Zhang Y, et al (2022) Motr: End-to-end multiple-object tracking with transformer. In: European conference on computer vision, Springer, pp 659--675

  55. [63]

    International journal of computer vision 129:3069--3087

    Zhang Y, Wang C, Wang X, et al (2021) Fairmot: On the fairness of detection and re-identification in multiple object tracking. International journal of computer vision 129:3069--3087

  56. [64]

    In: European conference on computer vision, Springer, pp 1--21

    Zhang Y, Sun P, Jiang Y, et al (2022) Bytetrack: Multi-object tracking by associating every detection box. In: European conference on computer vision, Springer, pp 1--21

  57. [65]

    Mechanical Systems and Signal Processing 213:111343

    Zhong S, Peng B, He J, et al (2024) Kalman filtering based on dynamic perception of measurement noise. Mechanical Systems and Signal Processing 213:111343

  58. [66]

    a henb \

    Zhou X, Koltun V, Kr \"a henb \"u hl P (2020) Tracking objects as points. In: European conference on computer vision, Springer, pp 474--490

  59. [67]

    IEEE transactions on pattern analysis and machine intelligence 44(11):7380--7399

    Zhu P, Wen L, Du D, et al (2021) Detection and tracking meet drones challenge. IEEE transactions on pattern analysis and machine intelligence 44(11):7380--7399

  60. [68]

    sn-basic.bst

    FUNCTION identify.basic.version "sn-basic.bst" " [2024/07/19 v1.1 bibliography style]" * top ENTRY address archive author booktitle chapter doi edition editor eid eprint howpublished institution journal key keywords month note number organization pages publisher school series ...

  61. [69]

    write newline

    " write newline "" before.all 'output.state := FUNCTION add.period duplicate empty 'skip "." * add.blank if FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap dupl...

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.