Pith. sign in

REVIEW 4 major objections 5 minor 5 cited by

Exploring Enhanced Contextual Information for Video-Level Object Tracking

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

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

Pith's one-line read This paper claims that video-level context carried in Mamba hidden states, fused into each backbone block by cross-attention, lets MCITrack surpass token-based video trackers and reach state-of-the-art accuracy.

desk verdict A capable, well-executed tracking paper that earns its SOTA claims on the evidence shown, but the unnamed threshold and interval T in the inference protocol plus ablation-on-LaSOT leave reproducibility and potential overfitting questions that a referee should press. read the letter →

arxiv 2412.11023 v1 pith:OFO3LU3K submitted 2024-12-15 cs.CV

Ben Kang , Xin Chen , Simiao Lai , Yang Liu , Yi Liu , Dong Wang This is my paper · ORCID
classification cs.CV
keywords videoobjecttrackingMambastatespacemodelhiddenstatescontextualinformationcross-attentionLaSOTGOT-10k
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 introduces MCITrack, a video-level object tracker that replaces the few extra tokens used by prior video-level trackers with Mamba hidden states that continuously record historical context. The central claim is that these hidden states transmit richer contextual information than a handful of tokens, and the numbers offered are state-of-the-art accuracy of 76.6% AUC on LaSOT and 80.0% average overlap on GOT-10k. A sympathetic reader would care because, if true, the bottleneck in video tracking shifts from how much context can be packed into extra tokens to how well a recurrent state can compress the video history.

What carries the argument

The Contextual Information Fusion (CIF) module is the load-bearing component, and its core is the Mamba layer, a selective state space model whose hidden state $h_t = \bar{A} h_{t-1} + \bar{B} x_t$ carries the video history. The hidden state is updated only when the tracker's classification score beats an unspecified threshold, at an update interval $T$, using a memory bank of reliable frames to avoid injecting erroneous context. Four CIF blocks are paired with four backbone blocks, so context is injected at multiple feature levels rather than once at the end. This machinery is what the paper claims transfers more contextual information than the few extra tokens of prior methods.

What would settle it

Run MCITrack-L384 on LaSOT with the hidden-state update threshold fixed to a constant (for example 0.5) and the update interval set to every frame, then compare the AUC to the reported 76.6%; a large drop would indicate the numbers depend on undisclosed update tuning rather than on hidden-state capacity.

Watch

Extended reading notes

Core claim

The paper claims that carrying video-level context in Mamba hidden states, fused into each backbone block through cross-attention, transmits more information than the additional tokens used by ODTrack and similar trackers. Its Contextual Information Fusion (CIF) module combines a Mamba layer that stores and updates a fixed-size hidden state with in-attention and out-attention layers that inject this context before each backbone block and extract current-frame features back into the state. On standard benchmarks, the largest variant MCITrack-L384 reports 76.6% AUC on LaSOT and 80.0% average overlap on GOT-10k, which the paper presents as the best published numbers on both.

Load-bearing premise

The benchmark gains assume that the hidden-state update rule (the classification-score threshold and the update interval $T$, whose values are never stated) was not tuned on the same LaSOT test set used for the headline result, and that this rule transfers to long videos without re-tuning.

Editorial extensions

If this is right

  • Video-level trackers no longer need to grow token counts to carry context; a fixed-size recurrent state can in principle serve the same role with less overhead.
  • Multi-level injection means context can help shallow high-resolution features and deep semantic features simultaneously, which may reduce drift in long-term tracking.
  • The reported efficiency (MCITrack-B224 beats ODTrack-B384 with lower FLOPs and a lower input resolution) suggests hidden-state context transmission is cheaper than token-based propagation.
  • The CIF module is a generic wrapper, so the same architectural recipe could be applied to other transformer backbones beyond the Fast-iTPN used here.

Reading between the lines

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

  • The hidden-state approach might transfer to other video-sequence tasks such as action recognition or multi-object tracking, where a compact stateful summary of history could replace token queues.
  • The undisclosed update threshold could likely be replaced by a learned confidence gate, removing a hyperparameter while retaining robustness to unreliable frames.
  • A stress test on very long videos (thousands of frames) would reveal whether the hidden state saturates or forgets early context, since the reported benchmarks are mostly on minute-scale sequences.
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 MCITrack, a video-level object tracker that propagates contextual information through Mamba hidden states instead of the few extra tokens used by prior video-level trackers. The core Contextual Information Fusion (CIF) module pairs each backbone block with a mamba layer and two cross-attention layers: the mamba layer stores historical context in hidden states, the in-attention layer injects that context into backbone features, and the out-attention layer extracts current-frame information back into the CIF block. The model is trained with 5-frame clips and a two-step unroll, then evaluated on eight tracking benchmarks. The central empirical claim is state-of-the-art performance, including 76.6% AUC on LaSOT and 80.0% AO on GOT-10k, with ablations on LaSOT supporting the CIF components, the clip length, the hidden-state size, and the choice of Mamba over extra tokens, LSTM, or previous-frame features.

Significance. If the reported results hold under a fixed and fully specified inference protocol, the paper makes a useful contribution to video-level tracking: the architecture is clearly described, the ablations cover the main design choices, the model family spans a reasonable efficiency-accuracy trade-off, and code and models are promised. The strength of the empirical case is reduced by two issues: the inference-time gating protocol has unspecified parameters that directly affect the reported numbers, and several key design choices were selected on the same LaSOT benchmark used for the headline state-of-the-art claim. These issues are addressable and do not invalidate the architectural idea, but they need to be resolved before the central claim can be fully trusted.

major comments (4)
  1. [Implementation Details, Inference] The inference protocol is incomplete: the paper says hidden states are updated only when the classification score exceeds a threshold and that the video clip is refreshed at an update interval T using frames from a memory bank, but it never gives numerical values for the threshold, T, or the memory-bank update rule. These parameters determine which frames propagate into the hidden state and therefore directly affect the reported AUC/AO numbers. Please specify the threshold, T, the memory-bank capacity and update rule, and state whether these values were chosen before or after seeing test-set results; without that information the method cannot be reproduced and comparisons with ODTrack and ARTrackV2 may reflect differences in inference heuristics rather than in the CIF architecture.
  2. [Implementation Details, Training vs. Inference] Training updates hidden states on every step with no confidence gating, while inference gates updates by the classification score and periodically refreshes the clip at interval T. This training-inference mismatch is not discussed. If the threshold and T were tuned on LaSOT test, the headline 76.6% AUC would be an upper envelope rather than the performance of a fixed method. Please clarify whether gating was applied during training, and if it was not, report the sensitivity of the main results to the threshold and T on a validation split, and confirm that all compared methods and all ablation variants use exactly the same inference protocol.
  3. [Tables 4-6 and State-of-the-Art Comparisons] The hidden-state size (16), the number of CIF blocks (4), the video clip length (5), and the propagation method are selected by ablations on LaSOT, and LaSOT is also the benchmark used for the headline 'unprecedented AUC' claim. Selecting the final configuration on the same test set used for the state-of-the-art comparison can inflate the reported gains over the 'Extra Token' and 'LSTM' alternatives. Please report ablations on a held-out split or on additional benchmarks, or otherwise justify that the final configuration was frozen before test evaluation.
  4. [Table 6, Contextual Information Propagation Methods] The ablation table does not state whether the Baseline, Extra Token, LSTM, and Previous Features rows all used the same hidden-state update threshold, interval T, and memory-bank protocol. If the Mamba baseline benefited from a tuned gating rule while the alternatives used a naive or fixed update, the +1.9% AUC over Extra Token would be attributable to the inference heuristic rather than to the hidden-state representation. Please specify the protocol used in each row and, ideally, include repeated-run variance or results on another benchmark, because the reported differences (0.9-1.9 AUC) are small and no significance information is provided.
minor comments (5)
  1. [Implementation Details, Inference] The text contains a typo: 'set a thresholda for updating' should read 'set a threshold for updating'.
  2. [Tables 1, 3, 4, 5, 6] The column header 'P N orm P' appears with broken spacing in several tables and should be rendered as 'P Norm P'.
  3. [Model Architecture, Backbone and State Space Model] The symbol N is used both for the number of backbone/CIF blocks and for the SSM state size in Eq. (2); renaming one of them would avoid ambiguity.
  4. [Implementation Details, Training] The training paragraph says a 5-frame video clip and two search regions are used, but the overview describes inputs as one video clip and one search region; please clarify how the two search regions are sampled and paired with the clip.
  5. [Limitation] The Limitation paragraph addresses training speed and clip overhead but does not mention the sensitivity of the results to the inference threshold and interval T; adding a sentence on the chosen values and their effect would strengthen reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: MCITrack is an empirical architecture paper whose benchmark claims are evaluated externally, not derived from its own inputs.

full rationale

This paper contains no formal derivation chain whose output is equivalent to its inputs. The central claims are empirical: a proposed architecture (Mamba hidden states plus cross-attention fusion) is trained and evaluated on standard tracking benchmarks. The headline LaSOT/GOT-10k numbers come from external test sets, not from the paper's equations. The ablation study in Table 6 compares Mamba with alternatives such as extra tokens and LSTM; this is a direct empirical comparison rather than a definitional identity. The unspecified inference threshold and update interval T are reproducibility and possibly overfitting concerns, but they are not shown to make any reported metric equal to a fitted parameter by construction. Hyperparameters such as hidden state size and clip length are selected via ablation on LaSOT, which may inflate the reported LaSOT number if the test set was used for selection, but this is an evaluation-protocol issue, not a circular reduction: the final AUC is not defined as the maximizer of the ablation grid. Self-citations appear only in related work and are not load-bearing. No uniqueness theorem, ansatz smuggling, or renaming of a known result is present. Under the stated standard for circularity, the paper is self-contained against external benchmarks.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

MCITrack introduces no new physical or conceptual entities in the sense of a new force, particle, or dimension. Its contributions are architectural: a Mamba hidden-state carrier, per-block cross-attention, and a selective update schedule. The free parameters are mostly standard training and architecture hyperparameters, but several are tuned on the same LaSOT benchmark that anchors the SOTA claim, which is the main circularity burden.

free parameters (6)
  • Hidden state size = 16
    Chosen as baseline after ablation on LaSOT; sizes 4, 8, and 32 all give lower AUC (Table 4).
  • Number of CIF blocks = 4
    Ablation shows 2 blocks drops AUC by 1.4 and 6 blocks by 0.5 on LaSOT (Table 4).
  • Video clip length = 5 frames
    Ablation on LaSOT: 2, 3, 4, and 6 frames all give lower AUC than 5 frames (Table 5).
  • Loss weights lambda_c, lambda_l, lambda_g = 1, 5, 2
    Hyperparameters in Eq. 5, styled after single-frame tracker losses; no task-specific derivation is given.
  • Hidden-state update threshold = not specified in text
    Inference updates hidden states only when classification score exceeds an unnamed threshold; the numeric value is not provided.
  • Memory bank update interval T = not specified in text
    Inference refreshes the video clip from the memory bank at interval T, but T is never given.
assumptions (5)
  • standard math Zero-order hold discretization of the SSM (Eq. 3) is valid.
    Mamba's discretized state transition A_bar = exp(Delta A) and B_bar approximately equal to Delta B is taken as given from Gu and Dao 2023.
  • domain assumption Mamba hidden states can store and transmit useful video context across frames.
    The central design uses h_t = A_bar h_{t-1} + B_bar x_t to carry information; its usefulness is asserted by the authors and supported only by ablation, not by an independent information-theoretic argument.
  • domain assumption Cross-attention between hidden states and backbone features at every block improves feature extraction.
    Ablations show removing in/out attention lowers AUC, but the mechanism is not proven beyond these single-dataset experiments.
  • ad hoc to paper The 5-frame clip and search-region patch concatenation form a valid sequence input.
    Training and inference rely on a 5-frame video clip plus current search region; this is a design choice validated only by LaSOT ablation.
  • domain assumption The selective hidden-state update is a safe inference strategy.
    Updating hidden states only when classification score exceeds an unspecified threshold is assumed to prevent error propagation, but no sensitivity analysis or threshold value is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Exploring Enhanced Contextual Information for Video-Level Object Tracking." pith.science (2026). https://pith.science/paper/OFO3LU3K

@misc{pith2026241211023,
  author       = {Pith},
  title        = {Pith review of: Exploring Enhanced Contextual Information for Video-Level Object Tracking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OFO3LU3K}},
  note         = {Machine review of arXiv:2412.11023}
}
read the original abstract

Contextual information at the video level has become increasingly crucial for visual object tracking. However, existing methods typically use only a few tokens to convey this information, which can lead to information loss and limit their ability to fully capture the context. To address this issue, we propose a new video-level visual object tracking framework called MCITrack. It leverages Mamba's hidden states to continuously record and transmit extensive contextual information throughout the video stream, resulting in more robust object tracking. The core component of MCITrack is the Contextual Information Fusion module, which consists of the mamba layer and the cross-attention layer. The mamba layer stores historical contextual information, while the cross-attention layer integrates this information into the current visual features of each backbone block. This module enhances the model's ability to capture and utilize contextual information at multiple levels through deep integration with the backbone. Experiments demonstrate that MCITrack achieves competitive performance across numerous benchmarks. For instance, it gets 76.6% AUC on LaSOT and 80.0% AO on GOT-10k, establishing a new state-of-the-art performance. Code and models are available at https://github.com/kangben258/MCITrack.

Figures

Figures reproduced from arXiv: 2412.11023 by the authors.

Figure 1
Figure 1. Comparison of different contextual information [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. (a) Framework of the proposed MCITrack. The key components include the backbone for visual feature extraction, [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The detailed mamba layer in the CIF block. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: EAO rank plots on VOT2020. shown in [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Visualization of the output features of CIF module. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. What You Have is What You Track: Adaptive and Robust Multimodal Tracking

    cs.CV 2025-07 conditional novelty 6.0 of 10

    FlexTrack claims state-of-the-art multimodal tracking on complete and simulated missing-modality benchmarks, using heterogeneous mixture-of-experts fusion and a video-level masking training strategy.

  2. AuroraLong: Bringing RNNs Back to Efficient Open-Ended Video Understanding

    cs.CV 2025-07 conditional novelty 6.0 of 10

    A 2B-parameter video-language model using an RWKV linear-RNN backbone and sorted token merging achieves competitive long-video QA accuracy with far lower memory cost than transformer-based models.

  3. SonarT165: A Large-scale Benchmark and STFTrack Framework for Acoustic Object Tracking

    cs.CV 2025-04 conditional novelty 6.0 of 10

    The paper introduces SonarT165, a large benchmark for underwater acoustic object tracking, and STFTrack, a tracker with multi-view template fusion and trajectory correction that reports state-of-the-art results on it.

  4. Towards Compact Unified Multimodal Tracking: Synergizing Knowledge Distillation with Structural Pruning

    cs.CV 2026-08 conditional novelty 4.0 of 10

    A pruned-head student with dual spatial and semantic distillation reaches 54 FPS and near-teacher accuracy on RGB-T and RGB-E tracking.

  5. FocusTrack: A Self-Adaptive Local Sampling Algorithm for Efficient Anti-UAV Tracking

    cs.CV 2025-04 conditional novelty 4.0 of 10

    FocusTrack adaptively enlarges a local tracker's search region when its confidence drops, and uses attention-generated masks to refine features, reaching 67.7% AUC on AntiUAV and 62.8% on AntiUAV410 at real-time speeds.

Reference graph

Works this paper leans on

69 extracted references · 60 canonical work pages · cited by 5 Pith papers

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bai, Y.; Zhao, Z.; Gong, Y.; and Wei, X. 2024. ARTrackV2: Prompting Autoregressive Tracker Where to Look and How to Describe. In CVPR, 19048--19057

    work page 2024

  4. [4]

    F.; Vedaldi, A.; and Torr, P

    Bertinetto, L.; Valmadre, J.; Henriques, J. F.; Vedaldi, A.; and Torr, P. H. S. 2016. Fully-Convolutional Siamese Networks for Object Tracking. In ECCV, 850--865

    work page 2016

  5. [5]

    V.; and Timofte, R

    Bhat, G.; Danelljan, M.; Gool, L. V.; and Timofte, R. 2019. Learning Discriminative Model Prediction for Tracking. In ICCV, 6182--6191

    work page 2019

  6. [6]

    Cai, Y.; Liu, J.; Tang, J.; and Wu, G. 2023. Robust Object Modeling for Visual Tracking. In ICCV, 9589--9600

    work page 2023

  7. [7]

    Cao, Z.; Huang, Z.; Pan, L.; Zhang, S.; Liu, Z.; and Fu, C. 2022. TCTrack: Temporal Contexts for Aerial Tracking. In CVPR, 14798--14808

    work page 2022

  8. [8]

    Chang, L.; Yongsheng, Y.; Xin, C.; Huchuan, L.; and Dong, W. 2024. Spatial-temporal initialization dilemma: towards realistic visual tracking. Visual Intelligence 2, Article no. 35

    work page 2024

Show all 69 references
  1. [9]

    Chen, B.; Li, P.; Bai, L.; Qiao, L.; Shen, Q.; Li, B.; Gan, W.; Wu, W.; and Ouyang, W. 2022. Backbone is All Your Need: A Simplified Architecture for Visual Object Tracking. In ECCV, 375--392

    2022

  2. [10]

    Chen, X.; Peng, H.; Wang, D.; Lu, H.; and Hu, H. 2023 a . SeqTrack: Sequence to Sequence Learning for Visual Object Tracking. In CVPR, 14572--14581

    2023

  3. [11]

    Chen, X.; Yan, B.; Zhu, J.; Lu, H.; Ruan, X.; and Wang, D. 2023 b . High-Performance Transformer Tracking. IEEE TPAMI, 8507--8523

    2023

  4. [12]

    Chen, X.; Yan, B.; Zhu, J.; Wang, D.; Yang, X.; and Lu, H. 2021. Transformer Tracking. In CVPR, 8126--8135

    2021

  5. [13]

    Chen, Z.; Zhong, B.; Li, G.; Zhang, S.; and Ji, R. 2020. Siamese Box Adaptive Network for Visual Tracking. In CVPR, 6668--6677

    2020

  6. [14]

    Cui, Y.; Jiang, C.; Wang, L.; and Wu, G. 2022. MixFormer: End-to-End Tracking with Iterative Mixed Attention. In CVPR, 13608--13618

    2022

  7. [15]

    Cui, Y.; Jiang, C.; Wang, L.; and Wu, G. 2024. MixFormer: End-to-End Tracking with Iterative Mixed Attention. IEEE TPAMI, 0--18

    2024

  8. [16]

    S.; and Felsberg, M

    Danelljan, M.; Bhat, G.; Khan, F. S.; and Felsberg, M. 2019. ATOM: Accurate Tracking by Overlap Maximization. In CVPR, 4660--4669

    2019

  9. [17]

    Dosovitskiy, A.; Beyer, L.; Kolesnikov, A.; Weissenborn, D.; Zhai, X.; Unterthiner, T.; Dehghani, M.; Minderer, M.; Heigold, G.; Gelly, S.; et al. 2020. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. In ICLR, 1--9

    2020

  10. [18]

    Elfwing, S.; Uchibe, E.; and Doya, K. 2018. Sigmoid-Weighted Linear Units for Neural Network Function Approximation in Reinforcement Learning. Neural Networks, 3--11

    2018

  11. [19]

    Fan, H.; Bai, H.; Lin, L.; Yang, F.; Chu, P.; Deng, G.; Yu, S.; Huang, M.; Liu, J.; Xu, Y.; et al. 2021. LaSOT: A High-Quality Large-Scale Single Object Tracking Benchmark. IJCV, 439--461

    2021

  12. [20]

    Fan, H.; Lin, L.; Yang, F.; Chu, P.; Deng, G.; Yu, S.; Bai, H.; Xu, Y.; Liao, C.; and Ling, H. 2019. LaSOT : A High-Quality Benchmark for Large-Scale Single Object Tracking. In CVPR, 5374--5383

    2019

  13. [21]

    Fu, Z.; Liu, Q.; Fu, Z.; and Wang, Y. 2021. STMTrack: Template-free Visual Tracking with Space-time Memory Networks. In CVPR, 13774--13783

    2021

  14. [22]

    Gao, S.; Zhou, C.; Ma, C.; Wang, X.; and Yuan, J. 2022. AiATrack : Attention in Attention for Transformer Visual Tracking. In ECCV, 146--164

    2022

  15. [23]

    Graves, A. 2012. Supervised Sequence Labelling with Recurrent Neural Networks. Studies in Computational Intelligence. Springer

    2012

  16. [24]

    Gu, A.; and Dao, T. 2023. Mamba: Linear-Time Sequence Modeling with Selective State Spaces. arXiv preprint arXiv:2312.00752

    2023 arXiv

  17. [25]

    Gu, A.; Goel, K.; and R \' e , C. 2022. Efficiently Modeling Long Sequences with Structured State Spaces. In ICLR, 1--9

    2022

  18. [26]

    Gu, A.; Johnson, I.; Goel, K.; Saab, K.; Dao, T.; Rudra, A.; and R \'e , C. 2021. Combining Recurrent, Convolutional, and Continuous-Time Models with Linear State Space Layers. In NeurIPS, 572--585

    2021

  19. [27]

    Guo, M.; Zhang, Z.; Fan, H.; Jing, L.; Lyu, Y.; Li, B.; and Hu, W. 2022. Learning Target-aware Representation for Visual Tracking via Informative Interactions. In IJCAI, 927--934

    2022

  20. [28]

    He, K.; Zhang, C.; Xie, S.; Li, Z.; and Wang, Z. 2023. Target-Aware Tracking with Long-Term Context Attention. In AAAI, 773--780

    2023

  21. [29]

    Huang, L.; Zhao, X.; and Huang, K. 2019. GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in The Wild. IEEE TPAMI, 1562--1577

    2019

  22. [30]

    Kiani Galoogahi, H.; Fagg, A.; Huang, C.; Ramanan, D.; and Lucey, S. 2017. Need for Speed: A Benchmark for Higher Frame Rate Object Tracking. In ICCV, 1125--1134

    2017

  23. [31]

    a m \"a r \

    Kristan, M.; Leonardis, A.; Matas, J.; Felsberg, M.; Pflugfelder, R.; K \"a m \"a r \"a inen, J.-K.; Danelljan, M.; Zajc, L. C .; Luke z i c , A.; Drbohlav, O.; et al. 2020. The Eighth Visual Object Tracking VOT 2020 Challenge Results. In ECCV, 547--601

    2020

  24. [32]

    Li, B.; Wu, W.; Wang, Q.; Zhang, F.; Xing, J.; and Yan, J. 2019. SiamRPN++ : Evolution of Siamese Visual Tracking with Very Deep Networks. In CVPR, 4282--4291

    2019

  25. [33]

    Li, B.; Yan, J.; Wu, W.; Zhu, Z.; and Hu, X. 2018. High Performance Visual Tracking With Siamese Region Proposal Network. In CVPR, 8971--8980

    2018

  26. [34]

    Li, X.; Huang, Y.; He, Z.; Wang, Y.; Lu, H.; and Yang, M.-H. 2023. Citetracker: Correlating Image and Text for Visual Tracking. In ICCV, 9974--9983

    2023

  27. [35]

    Lin, L.; Fan, H.; Xu, Y.; and Ling, H. 2022. Swintrack: A Simple and Strong Baseline for Transformer Tracking. In NeurIPS, 16743--16754

    2022

  28. [36]

    Lin, L.; Fan, H.; Zhang, Z.; Wang, Y.; Xu, Y.; and Ling, H. 2024. Tracking Meets LoRA: Faster Training, Larger Model, Stronger Performance. In ECCV, 1--15

    2024

  29. [37]

    Lin, T.-Y.; Goyal, P.; Girshick, R.; He, K.; and Doll \'a r, P. 2017. Focal Loss for Dense Object Detection. In ICCV, 2980--2988

    2017

  30. [38]

    J.; Bourdev, L

    Lin, T.-Y.; Maire, M.; Belongie, S. J.; Bourdev, L. D.; Girshick, R. B.; Hays, J.; Perona, P.; Ramanan, D.; Doll \'a r, P.; and Zitnick, C. L. 2014. Microsoft COCO : Common Objects in Context. In ECCV, 740--755

    2014

  31. [39]

    Liu, Y.; Tian, Y.; Zhao, Y.; Yu, H.; Xie, L.; Wang, Y.; Ye, Q.; and Liu, Y. 2024. VMamba: Visual State Space Model. arXiv preprint arXiv:2401.10166

    2024 arXiv

  32. [40]

    Loshchilov, I.; and Hutter, F. 2018. Decoupled Weight Decay Regularization. In ICLR, 1--9

    2018

  33. [41]

    P.; Yu, F.; and Van Gool, L

    Mayer, C.; Danelljan, M.; Bhat, G.; Paul, M.; Paudel, D. P.; Yu, F.; and Van Gool, L. 2022. Transforming Model Prediction for Tracking. In CVPR, 8731--8740

    2022

  34. [42]

    P.; and Van Gool, L

    Mayer, C.; Danelljan, M.; Paudel, D. P.; and Van Gool, L. 2021. Learning Target Candidate Association to Keep Track of What Not to Track. In ICCV, 13444--13454

    2021

  35. [43]

    Mueller, M.; Smith, N.; and Ghanem, B. 2016. A Benchmark and Simulator for UAV Tracking. In ECCV, 445--461

    2016

  36. [44]

    Muller, M.; Bibi, A.; Giancola, S.; Alsubaihi, S.; and Ghanem, B. 2018. Tracking N et: A Large-Scale Dataset and Benchmark for Object Tracking in The Wild. In ECCV, 300--317

    2018

  37. [45]

    Peng, L.; Gao, J.; Liu, X.; Li, W.; Dong, S.; Zhang, Z.; Fan, H.; and Zhang, L. 2024. VastTrack: Vast Category Visual Object Tracking. arXiv preprint arXiv:2403.03493

    2024 arXiv

  38. [46]

    D.; and Savarese, S

    Rezatofighi, H.; Tsoi, N.; Gwak, J.; Sadeghian, A.; Reid, I. D.; and Savarese, S. 2019. Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression. In CVPR, 658--666

    2019

  39. [47]

    Ronneberger, O.; Fischer, P.; and Brox, T. 2015. U-Net: Convolutional Networks for Biomedical Image Segmentation. In MICCAI, 234--241

    2015

  40. [48]

    Shi, L.; Zhong, B.; Liang, Q.; Li, N.; Zhang, S.; and Li, X. 2024. Explicit Visual Prompts for Visual Object Tracking. In AAAI, 4838--4846

    2024

  41. [49]

    P.; and Yang, W

    Song, Z.; Luo, R.; Yu, J.; Chen, Y.-P. P.; and Yang, W. 2023. Compact Transformer Tracker with Correlative Masked Modeling. In AAAI, 2321--2329

    2023

  42. [50]

    Tao, R.; Gavves, E.; and Smeulders, A. W. M. 2016. Siamese Instance Search for Tracking. In CVPR, 1420--1429

    2016

  43. [51]

    Tian, Y.; Xie, L.; Qiu, J.; Jiao, J.; Wang, Y.; Tian, Q.; and Ye, Q. 2024. Fast-iTPN: Integrally Pre-Trained Transformer Pyramid Network with Token Migration. IEEE TPAMI, 1--15

    2024

  44. [52]

    Voigtlaender, P.; Luiten, J.; Torr, P. H. S.; and Leibe, B. 2020. Siam R-CNN: V isual Tracking by Re-Detection. In CVPR, 6578--6588

    2020

  45. [53]

    Wang, N.; Zhou, W.; Wang, J.; and Li, H. 2021 a . Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Tracking. In CVPR, 1571--1580

    2021

  46. [54]

    Wang, X.; Shu, X.; Zhang, Z.; Jiang, B.; Wang, Y.; Tian, Y.; and Wu, F. 2021 b . Towards More Flexible and Accurate Object Tracking with Natural Language: Algorithms and Benchmark. In CVPR, 13763--13773

    2021

  47. [55]

    Wang, Z.; Zheng, J.-Q.; Zhang, Y.; Cui, G.; and Li, L. 2024. Mamba-Unet: Unet-Like Pure Visual Mamba for Medical Image Segmentation. arXiv preprint arXiv:2402.05079

    2024 arXiv

  48. [56]

    Wei, X.; Bai, Y.; Zheng, Y.; Shi, D.; and Gong, Y. 2023. Autoregressive Visual Tracking. In CVPR, 9697--9706

    2023

  49. [57]

    Xie, F.; Chu, L.; Li, J.; Lu, Y.; and Ma, C. 2023. VideoTrack: Learning to Track Objects via Video Transformer. In CVPR, 22826--22835

    2023

  50. [58]

    Xie, F.; Wang, C.; Wang, G.; Cao, Y.; Yang, W.; and Zeng, W. 2022. Correlation-Aware Deep Tracking. In CVPR, 8751--8760

    2022

  51. [59]

    Xie, J.; Zhong, B.; Mo, Z.; Zhang, S.; Shi, L.; Song, S.; and Ji, R. 2024. Autoregressive Queries for Adaptive Tracking with Spatio-Temporal Transformers. In CVPR, 19300--19309

    2024

  52. [60]

    Xu, Y.; Wang, Z.; Li, Z.; Yuan, Y.; and Yu, G. 2020. SiamFC++: T owards Robust and Accurate Visual Tracking with Target Estimation Guidelines. In AAAI, 12549--12556

    2020

  53. [61]

    Yan, B.; Peng, H.; Fu, J.; Wang, D.; and Lu, H. 2021 a . Learning Spatio-Temporal Transformer for Visual Tracking. In ICCV, 10448--10457

    2021

  54. [62]

    Yan, B.; Zhang, X.; Wang, D.; Lu, H.; and Yang, X. 2021 b . Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation. In CVPR, 5289--5298

    2021

  55. [63]

    Ye, B.; Chang, H.; Ma, B.; Shan, S.; and Chen, X. 2022. Joint Feature Learning and Relation Modeling for Tracking: A One-Stream Framework. In ECCV, 341--357

    2022

  56. [64]

    Zhang, B.; and Sennrich, R. 2019. Root Mean Square Layer Normalization. In NeurIPS, 12360--12371

    2019

  57. [65]

    Zhang, Z.; and Peng, H. 2019. Deeper and Wider Siamese Networks for Real-Time Visual Tracking. In CVPR, 4591--4600

    2019

  58. [66]

    Zhang, Z.; Peng, H.; Fu, J.; Li, B.; and Hu, W. 2020. Ocean: Object-aware Anchor-free Tracking. In ECCV, 771--787

    2020

  59. [67]

    Zheng, Y.; Zhong, B.; Liang, Q.; Mo, Z.; Zhang, S.; and Li, X. 2024. ODTrack: Online Dense Temporal Token Learning for Visual Tracking. In AAAI, 7588--7596

    2024

  60. [68]

    Zhu, J.; Lai, S.; Chen, X.; Wang, D.; and Lu, H. 2023. Visual prompt multi-modal tracking. In CVPR, 9516--9526

    2023

  61. [69]

    Zhu, L.; Liao, B.; Zhang, Q.; Wang, X.; Liu, W.; and Wang, X. 2024. Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model. arXiv preprint arXiv:2401.09417

    2024 arXiv

Pith tools

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