Pith. sign in

REVIEW 4 major objections 5 minor 67 references

Improving Accuracy and Generalization for Efficient Visual Tracking

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read The paper claims that a ~2M-parameter Siamese tracker, SiamABC, tracks objects under out-of-distribution conditions at 100 FPS on a CPU, beating MixFormerV2-S by 7.6% AUC on AVisT while running almost 3x faster.

desk verdict A solid empirical tracking paper whose headline result depends on an underspecified test-time update that the printed Algorithm 2 cannot actually execute; worth reviewing if fixed. read the letter →

arxiv 2411.18855 v2 pith:GKTPJXXB submitted 2024-11-28 cs.CV cs.LGcs.MM

classification cs.CVcs.LGcs.MM
keywords visualobjecttrackingSiamesetrackerefficientinferenceout-of-distributiongeneralizationtest-timeadaptationbatchnormalizationstatisticstransitiverelationlossFastMixedFiltration
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 a lightweight Siamese tracker can track objects on out-of-distribution (OOD) video, footage unlike its training data, about as accurately as much larger models while running at 100 frames per second on a CPU. The method, SiamABC, adds a dual-search-region to keep the tracker anchored to both the initial target and its recent appearance, a learnable Fast Mixed Filtration layer to combine template and search features cheaply, a transitive relation loss for training, and a backward-free dynamic test-time adaptation that updates batch-normalization statistics during inference. On the AVisT adverse-conditions benchmark, the smallest model, S-Tiny, reaches 47.2% AUC, outperforming MixFormerV2-S by 7.6% while being almost three times faster. The paper also claims this is the first use of test-time adaptation for single-object visual tracking. If the results hold, efficient trackers could be deployed in-the-wild under constrained resources with less accuracy loss than current efficient methods.

What carries the argument

The central object is Fast Mixed Filtration (FMF), a compact channel and spatial attention block built from 1x1 convolutions, softmax masks, and broadcast element-wise multiplications and additions. FMF filters the concatenated dual-template and dual-search-region features and compresses them from 2C channels back to C; the paper reports 0.034 GFLOPs, 0.395M parameters, and 0.4ms CPU latency, matching the accuracy of polarized self-attention at half the latency. Two other load-bearing pieces are the transitive relation loss, which pulls the filtered template and search-region representations close in cosine space using a projection head with a stop-gradient, and Dynamic Test-Time Adaptation (DTTA), a backward-free update of the head batch-normalization statistics kept anchored to the source statistics with a small anchor weight of 0.1. Together these carry the paper's accuracy and generalization claims.

What would settle it

Remove DTTA while keeping FMF, the transitive relation loss, and the dual-search-region, and measure the AUC gap on AVisT and ITB. The paper's own tables show a gap of about 1.4 AUC on AVisT and 0.9 on ITB; if a reimplementation or a larger benchmark sweep finds that gap is within benchmark noise, the BN-statistic premise is not carrying the OOD claim. A more direct test is to take a tracker trained on clean weather data, evaluate on heavily corrupted variants where batch-normalization statistics are the only shifted component, and check whether the anchored instance-statistic update recovers the lost accuracy.

Watch

Extended reading notes

Core claim

The central claim is that the speed-accuracy trade-off of efficient visual tracking can be pushed by bridging temporal context at both the template and the search-region ends, and by adapting the model at inference time without backpropagation. SiamABC maintains four inputs: a static template, a dynamic template, the current search region, and a dynamic search region re-centered on the object. Their features are filtered and correlated, and training is guided by a transitive relation loss. During inference, the batch-normalization statistics of the prediction heads are updated with instance statistics anchored to the source statistics. On AVisT, S-Tiny reaches 47.2% AUC at 100 FPS on a CPU, outperforming MixFormerV2-S, which gets 39.6% AUC at 37 FPS, by 7.6% while running almost three times faster.

Load-bearing premise

The load-bearing premise is that shifts in batch-normalization statistics are a major source of out-of-distribution accuracy loss in tracking; if that premise is false for tracking, the dynamic test-time adaptation contributes little beyond the architectural and loss improvements.

Editorial extensions

If this is right

  • S-Tiny reaches 47.2% AUC on AVisT, ahead of MixFormerV2-S by 7.6% while running at 100 FPS instead of 37 FPS on a CPU.
  • The FMF layer matches polarized self-attention accuracy with about half the CPU latency (0.4ms versus 0.8ms) and lower FLOPs.
  • Adding the transitive relation loss raises AVisT AUC from 43.7% to 45.8% in the ablation, and the dynamic test-time adaptation lifts it further to 47.2%.
  • Across the 11 evaluated benchmarks the largest gains are on out-of-distribution sets such as AVisT, UAV123, ITB, and NFS30, while in-distribution performance stays competitive with efficient trackers.
  • DTTA is backward-free and adds negligible latency (3.7ms versus 3.6ms on a CPU), so adaptation remains compatible with real-time tracking.

Reading between the lines

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

  • Because DTTA only updates batch-normalization running statistics and leaves weights frozen, the same anchored-update recipe could be bolted onto other already-trained trackers at inference time, and possibly onto other single-frame video tasks, without retraining.
  • A controlled study that keeps the dual-search-region but removes the template update could separate the contribution of the extra temporal anchor from the contribution of simply seeing more examples.
  • The ablation numbers imply the OOD improvement is spread across components: FMF and the transitive relation loss add about 4 AUC points on AVisT, while DTTA adds about 1.4; a reimplementation on different corruption types would show whether the BN-statistic premise transfers beyond the tested benchmarks.
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 introduces SiamABC, a Siamese visual tracker with four claimed contributions: a dual-search-region representation, a lightweight Fast Mixed Filtration (FMF) layer, a transitive relation loss (TRL), and a backward-free dynamic test-time adaptation (DTTA) method. The central claim is that the smallest model, S-Tiny, achieves a 47.2% AUC on the OOD AVisT benchmark at 100 FPS on a CPU, outperforming MixFormerV2-S by 7.6% while being about 3x faster. The manuscript reports extensive experiments on 11 benchmarks, TTA baseline comparisons, and component ablations, with code and models made publicly available.

Significance. If the reported results hold, the paper addresses a real gap: efficient trackers that are deployable on CPU while generalizing to out-of-distribution sequences. The architectural components are coherent, and the empirical scope is broad, including 11 benchmarks and latency measurements on CPU, GPU, and an edge device. The speed-accuracy trade-off claimed for S-Tiny is attractive and largely consistent with the presented tables. The ablation study is structured and the authors provide code and models, which are strengths. The main reservations are that the printed inference-time update algorithm is not implementable as written, the 'parameter-free' characterization is contradicted by the required hyperparameters, and the OOD gains attributed to DTTA and the dynamic update are small and reported without error bars.

major comments (4)
  1. [Section 3.2, Eq. (4), Appendix A.3 Algorithm 2] The printed dynamic-update procedure is not executable. In Algorithm 2, the condition 'if C >= N & rho_t > rho_t' compares the current classification score with itself, so a reader implementing the algorithm would never update I_S and I_D. Eq. (4) defines a running average but Algorithm 2 neither maintains nor compares against it; the line 'rho_t = (1 - lambda_D) rho_t + lambda_D rho_t' is also self-referential. A faithful implementation would revert to the 'No Updates' row of Table 6 (0.448 AUC), losing the +1.0 AUC attributed to this component and, together with DTTA, 2.4 AUC of the headline 0.472. Please rewrite Eq. (4) and Algorithm 2 with distinct symbols for the current score, the previous running average, and the updated running average, and state the exact configuration used for Tables 3 and 5.
  2. [Section 3.3, Eq. (7)] The DTTA recurrence in Eq. (7) is ambiguous because the same symbols 'mu_{I,t}' and 'sigma^2_{I,t}' appear on both sides, while the text defines the right-hand side as the current instance statistics. The updated running statistics should be denoted with new symbols, and the paper should state the initialization, whether the update is applied per frame or only at update intervals, and how DTTA interacts with Algorithm 2. This matters because Table 5 credits DTTA with +1.4 AUC on AVisT (0.458 to 0.472), and the current text does not fully determine the implemented rule.
  3. [Section 3.2, Table 6, Section 3.3] The update strategy is repeatedly described as 'parameter-free' and 'parameterless,' but it depends on N=60 and lambda_D=0.25, and DTTA depends on lambda_BN=0.1; the training sampler also depends on Delta=150. No sensitivity analysis or validation-based selection procedure is reported for any of these values. Since these hyperparameters directly determine the OOD improvements shown in Tables 5 and 6, the 'parameter-free' claim is not supported, and the headline numbers are not tied to a reproducible selection rule. Please add sensitivity sweeps for N, lambda_D, and lambda_BN, or report a validation protocol.
  4. [Tables 3-6, Figure 4] All results are single point estimates without error bars or significance testing. The main OOD improvements are small in absolute terms (DTTA +1.4 AUC on AVisT, dynamic update +1.0 AUC, S-Tiny vs. SMAT +2.5 AUC), and tracking benchmarks have substantial per-sequence variance. Please report variance over sequences or bootstrap confidence intervals for the central speed-accuracy and OOD-generalization claims so that the reader can assess whether the differences are statistically meaningful.
minor comments (5)
  1. [Section 3, Figure 2] The method is called 'Fast Mixed Filtration (FMT)' in the overview but 'FMF' elsewhere; please unify the abbreviation consistently.
  2. [Table 3] The benchmark name 'UA V123' should be 'UAV123', and 'A VisT' should be typeset consistently (e.g., 'AVisT') throughout the tables and text.
  3. [Eq. (1)] The symbol 'P' is used for element-wise summation in the equation but resembles a parameter symbol; please use a distinct notation, such as a circled plus or an explicit 'sum', to avoid ambiguity.
  4. [Section 4.1] There is a typo in 'averging in about 441 frames per sequence'; it should read 'averaging about 441 frames per sequence'.
  5. [Algorithm 2] The comments in Algorithm 2 call N and lambda_D hyperparameters, while the surrounding text calls the strategy parameterless; please reconcile the terminology.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: benchmark results are externally evaluated, components are ablated, and no load-bearing self-citation or fitted-input-called-prediction step can be exhibited.

full rationale

The paper's central claims are evaluated on external benchmarks (AVisT, LaSOT, GOT-10k, TrackingNet, ITB, etc.), and the reported comparisons are against published trackers, not against quantities fitted from the same data. The novel components (dual-search-region, FMF, TRL, DTTA) are each tested through ablations in Figure 4 and Tables 1, 5, and 6, with DTTA compared to external TTA baselines (TENT, ETA, Momentum, DUA, IN, AdaBN) rather than derived from a theorem quoted from the authors' own prior work. The reference list contains no prior publication by the present authors that is used to justify a premise, so no uniqueness-imported-from-authors or ansatz-smuggled-in-via-citation pattern is present. The notation problems in Eq. (7) (same symbol for the updated statistic and the current instance statistic) and Algorithm 2 (the condition 'ρt > ρt' compares the classification score with itself, and no separate running-average variable is maintained) are genuine reproducibility and specification bugs, but they are under-specification, not circular reasoning: they do not make an output equal to an input by construction, and the headline numbers do not reduce to those equations alone. Similarly, hyperparameters such as lambda_BN, lambda_D, N, S, and Delta are set without a reported validation-based selection procedure, which is a tuning-transparency concern, but there is no evidence that a benchmark result is being renamed as a prediction. The borrowed premise from classification TTA literature (that BN-statistic shifts drive OOD degradation) is external support, not a self-derived first-principles result. Therefore, no circular step can be exhibited with the required specificity, and the honest finding is no significant circularity.

Assumptions & free parameters 6 free parameters · 4 assumptions · 3 invented entities

The paper's gains come from a set of architectural modules and test-time heuristics. The main external anchors are the OOD/ID benchmarks; the main internal choices are the hyperparameters listed above, several of which affect the reported margins.

free parameters (6)
  • lambda_BN (DTTA anchor strength) = 0.1
    Weight blending instance statistics with source BN statistics in Eq. 7. Chosen without a reported sensitivity analysis.
  • lambda_D (dynamic update momentum) = 0.25
    Momentum for the running-average classification score in Eq. 4 that gates template refresh. This is a hyperparameter despite the paper calling the strategy parameter-free.
  • N (dynamic update interval) = 60 frames
    Counter threshold before a score-based update decision. Tuned value; no sensitivity analysis is reported.
  • S (FMF squeeze rate) = 2
    Squeeze rate of the Fast Mixed Filtration block. Ablation in Figure 4 bottom shows S=2 beats S=1 on AVisT, so the choice affects the headline result.
  • Delta (training sampling interval) = 150
    Gap between template and search frames in training (Algorithm 1). The paper says it was set arbitrarily, yet it determines how far apart the dynamic and current frames are.
  • Loss weights lambda_FL, lambda_TR, lambda_Reg = 1, 1/3, 1/3
    Weights of the total tracking loss in Eq. 3. Set by hand, standard for tracking losses.
assumptions (4)
  • domain assumption BN statistics drift is a major cause of OOD performance degradation in tracking
    Invoked in Section 3.3 to justify DTTA, borrowing from classification TTA literature [33,41,44,46,49,52]. Not independently validated for tracking.
  • domain assumption The dual-search-region re-centered at the object provides useful temporal context
    Central architectural premise (Section 3). Supported only by ablations on external benchmarks, not by a mechanism analysis.
  • domain assumption Stop-gradient prevents collapse of the transitive relation loss
    The TRL in Eq. 2 relies on the SimSiam stop-gradient trick [7]; the paper does not analyze whether the loss could still collapse onto the dynamic components.
  • domain assumption Running-average classification score is a reliable confidence signal
    The dynamic update rule (Algorithm 2) updates the template only when rho_t > rho_t with a running average. In the presence of distractors, high scores may occur for wrong objects, which the paper does not analyze.
invented entities (3)
  • Dual-search-region representation (I_S, F_S) independent evidence
    purpose: Provides temporal context from the previous search region re-centered at the object, complementing the static template.
    Ablation in Figure 4 top shows adding the dynamic search region improves AVisT AUC from 41.7 to 41.8 (mix+concat) and with FMF to 43.7; Table 6 shows further gains from the update strategy.
  • Fast Mixed Filtration (FMF) layer independent evidence
    purpose: Efficient channel-spatial filtration of concatenated dual representations.
    Table 1 shows FMF matches PSA accuracy with 0.4 ms CPU latency; Figure 4 shows FMF raises AVisT AUC by about 2 points over naive concat; Figure 6 shows qualitative stability.
  • Dynamic Test-Time Adaptation (DTTA) update rule independent evidence
    purpose: Backward-free update of head BN statistics anchored to source statistics during inference.
    Table 5 shows DTTA improves AVisT AUC from 0.458 to 0.472 and ITB from 0.539 to 0.548 over the no-TTA baseline, with minimal latency overhead.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving Accuracy and Generalization for Efficient Visual Tracking." pith.science (2026). https://pith.science/paper/GKTPJXXB

@misc{pith2026241118855,
  author       = {Pith},
  title        = {Pith review of: Improving Accuracy and Generalization for Efficient Visual Tracking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GKTPJXXB}},
  note         = {Machine review of arXiv:2411.18855}
}
read the original abstract

Efficient visual trackers overfit to their training distributions and lack generalization abilities, resulting in them performing well on their respective in-distribution (ID) test sets and not as well on out-of-distribution (OOD) sequences, imposing limitations to their deployment in-the-wild under constrained resources. We introduce SiamABC, a highly efficient Siamese tracker that significantly improves tracking performance, even on OOD sequences. SiamABC takes advantage of new architectural designs in the way it bridges the dynamic variability of the target, and of new losses for training. Also, it directly addresses OOD tracking generalization by including a fast backward-free dynamic test-time adaptation method that continuously adapts the model according to the dynamic visual changes of the target. Our extensive experiments suggest that SiamABC shows remarkable performance gains in OOD sets while maintaining accurate performance on the ID benchmarks. SiamABC outperforms MixFormerV2-S by 7.6\% on the OOD AVisT benchmark while being 3x faster (100 FPS) on a CPU. Our code and models are available at https://wvuvl.github.io/SiamABC/.

Figures

Figures reproduced from arXiv: 2411.18855 by the authors.

Figure 1
Figure 1. Comparison of our trackers with others on the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overall Architecture. The Feature Extraction Block uses a readily available backbone to process the frames. The Relation￾Aware Block exploits representational relations among the dual-template and dual-search-region through our losses, LT R and LReg, where dual-template and dual-search-region representations are obtained via our learnable FMF layer. The Heads Block learns lightweight convolution layers to infer the … view at source ↗
Figure 3
Figure 3. Fast Mixed Filtration. This block serves as a lightweight and effective attention mechanism. The input x is fil￾tered to produce the compressed representations ˇx. The broadcast and element-wise operations make this block efficient on CPU. matrix multiplications causing the latency on CPU to still be considerably high. This motivated the development of our Fast Mixed Filtration (FMF), a new and more efficient mixed … view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Ablation study on the components of SiamABC-Tiny. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison on the AVisT [45] dataset with other efficient trackers, and with the further inclusion of Ocean. Under adverse visibility conditions, our tracker, S-Tiny, is relatively stable compared to the others while running at 100 FPS on a CPU. 14 [PITH_F…
Figure 6
Figure 6. Figure 6: Qualitative results on the ablation study of the FMF layer on the AVisT [ [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

67 extracted references · 50 canonical work pages

  1. [1]

    https://github.com/got- 10k/ toolkit

    GitHub - got-10k/toolkit: Official Python toolkit for generic object tracking benchmark GOT-10k and beyond — github.com. https://github.com/got- 10k/ toolkit. 12

  2. [2]

    Revisiting test time adaptation under online evaluation

    Motasem Alfarra, Hani Itani, Alejandro Pardo, Shyma Al- huwaider, Merey Ramazanova, Juan C P ´erez, Zhipeng Cai, Matthias M ¨uller, and Bernard Ghanem. Revisiting test time adaptation under online evaluation. arXiv preprint arXiv:2304.04795, 2023. 2, 13

  3. [3]

    Fully-convolutional siamese networks for object tracking

    Luca Bertinetto, Jack Valmadre, Joao F Henriques, Andrea Vedaldi, and Philip HS Torr. Fully-convolutional siamese networks for object tracking. In Computer Vision–ECCV 2016 Workshops: Amsterdam, The Netherlands, October 8- 10 and 15-16, 2016, Proceedings, Part II 14, pages 850–865. Springer, 2016. 2, 3, 6

  4. [4]

    Learning discriminative model prediction for track- ing

    Goutam Bhat, Martin Danelljan, Luc Van Gool, and Radu Timofte. Learning discriminative model prediction for track- ing. In Proceedings of the IEEE/CVF international confer- ence on computer vision, pages 6182–6191, 2019. 6, 7

  5. [5]

    Efficient visual tracking with exemplar trans- formers

    Philippe Blatter, Menelaos Kanakis, Martin Danelljan, and Luc Van Gool. Efficient visual tracking with exemplar trans- formers. In Proceedings of the IEEE/CVF Winter Confer- ence on Applications of Computer Vision, pages 1571–1581,

  6. [6]

    Fear: Fast, efficient, accu- rate and robust visual tracker

    Vasyl Borsuk, Roman Vei, Orest Kupyn, Tetiana Martyniuk, Igor Krashenyi, and Ji ˇri Matas. Fear: Fast, efficient, accu- rate and robust visual tracker. In European Conference on Computer Vision, pages 644–663. Springer, 2022. 1, 2, 4, 6, 8

  7. [7]

    Exploring simple siamese rep- resentation learning

    Xinlei Chen and Kaiming He. Exploring simple siamese rep- resentation learning. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 15750–15758, 2021. 4

  8. [8]

    Efficient visual tracking via hierarchical cross- attention transformer

    Xin Chen, Ben Kang, Dong Wang, Dongdong Li, and Huchuan Lu. Efficient visual tracking via hierarchical cross- attention transformer. In European Conference on Computer Vision, pages 461–477. Springer, 2022. 2, 6

Show all 67 references
  1. [9]

    Seqtrack: Sequence to sequence learning for visual ob- ject tracking

    Xin Chen, Houwen Peng, Dong Wang, Huchuan Lu, and Han Hu. Seqtrack: Sequence to sequence learning for visual ob- ject tracking. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14572– 14581, 2023. 1, 2, 6

  2. [10]

    Transformer tracking

    Xin Chen, Bin Yan, Jiawen Zhu, Dong Wang, Xiaoyun Yang, and Huchuan Lu. Transformer tracking. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8126–8135, 2021. 6, 7

  3. [11]

    Siamese box adaptive network for visual tracking

    Zedu Chen, Bineng Zhong, Guorong Li, Shengping Zhang, and Rongrong Ji. Siamese box adaptive network for visual tracking. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6668–6677,

  4. [12]

    Mixformerv2: Efficient fully transformer tracking.Advances in Neural Information Processing Systems, 36, 2024

    Yutao Cui, Tianhui Song, Gangshan Wu, and Limin Wang. Mixformerv2: Efficient fully transformer tracking.Advances in Neural Information Processing Systems, 36, 2024. 1, 2, 6, 7

  5. [13]

    Atom: Accurate tracking by overlap max- imization

    Martin Danelljan, Goutam Bhat, Fahad Shahbaz Khan, and Michael Felsberg. Atom: Accurate tracking by overlap max- imization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4660–4669,

  6. [14]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009. 3

  7. [15]

    Lasot: A high-quality large-scale single object track- ing benchmark

    Heng Fan, Hexin Bai, Liting Lin, Fan Yang, Peng Chu, Ge Deng, Sijia Yu, Harshit, Mingzhen Huang, Juehuan Liu, et al. Lasot: A high-quality large-scale single object track- ing benchmark. International Journal of Computer Vision , 129:439–461, 2021. 5, 6, 7, 12

  8. [16]

    Dual attention network for scene seg- mentation

    Jun Fu, Jing Liu, Haijie Tian, Yong Li, Yongjun Bao, Zhiwei Fang, and Hanqing Lu. Dual attention network for scene seg- mentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3146–3154,

  9. [17]

    Aiatrack: Attention in attention for trans- former visual tracking

    Shenyuan Gao, Chunluan Zhou, Chao Ma, Xinggang Wang, and Junsong Yuan. Aiatrack: Attention in attention for trans- former visual tracking. In European Conference on Com- puter Vision, pages 146–164. Springer, 2022. 1, 2

  10. [18]

    Separable self and mixed attention transformers for efficient object track- ing

    Goutam Yelluru Gopal and Maria A Amer. Separable self and mixed attention transformers for efficient object track- ing. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 6708–6717, 2024. 2, 6, 7

  11. [19]

    Target-aware tracking with long-term context attention

    Kaijie He, Canlong Zhang, Sheng Xie, Zhixin Li, and Zhi- wen Wang. Target-aware tracking with long-term context attention. arXiv preprint arXiv:2302.13840, 2023. 1, 8

  12. [20]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 3

  13. [21]

    Learning to track at 100 fps with deep regression networks

    David Held, Sebastian Thrun, and Silvio Savarese. Learning to track at 100 fps with deep regression networks. In Com- puter Vision–ECCV 2016: 14th European Conference, Am- sterdam, The Netherlands, October 11–14, 2016, Proceed- ings, Part I 14, pages 749–765. Springer, 2016. 2

  14. [22]

    Got-10k: A large high-diversity benchmark for generic object tracking in the wild

    Lianghua Huang, Xin Zhao, and Kaiqi Huang. Got-10k: A large high-diversity benchmark for generic object tracking in the wild. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 43(5):1562–1577, 2021. 1, 6, 7, 12, 13

  15. [23]

    Squeezenet: Alexnet-level accuracy with 50x fewer pa- rameters and¡ 0.5 mb model size

    Forrest N Iandola, Song Han, Matthew W Moskewicz, Khalid Ashraf, William J Dally, and Kurt Keutzer. Squeezenet: Alexnet-level accuracy with 50x fewer pa- rameters and¡ 0.5 mb model size. arXiv preprint arXiv:1602.07360, 2016. 2, 4

  16. [24]

    Exploring lightweight hierarchical vision transformers for efficient visual tracking

    Ben Kang, Xin Chen, Dong Wang, Houwen Peng, and Huchuan Lu. Exploring lightweight hierarchical vision transformers for efficient visual tracking. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 9612–9621, 2023. 2, 6

  17. [25]

    Need for speed: A benchmark for higher frame rate object tracking

    Hamed Kiani Galoogahi, Ashton Fagg, Chen Huang, Deva Ramanan, and Simon Lucey. Need for speed: A benchmark for higher frame rate object tracking. In Proceedings of the 9 IEEE International Conference on Computer Vision , pages 1125–1134, 2017. 6, 7

  18. [26]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 ,

  19. [27]

    Siamese neural networks for one-shot image recognition

    Gregory Koch, Richard Zemel, Ruslan Salakhutdinov, et al. Siamese neural networks for one-shot image recognition. In ICML deep learning workshop, volume 2. Lille, 2015. 2

  20. [28]

    The eighth visual object tracking vot2020 chal- lenge results

    Matej Kristan, Ale ˇs Leonardis, Ji ˇr´ı Matas, Michael Fels- berg, Roman Pflugfelder, Joni-Kristian K ¨am¨ar¨ainen, Martin Danelljan, Luka ˇCehovin Zajc, Alan Luke ˇziˇc, Ondrej Dr- bohlav, et al. The eighth visual object tracking vot2020 chal- lenge results. In Computer Visi...

  21. [29]

    Siamrpn++: Evolution of siamese vi- sual tracking with very deep networks

    Bo Li, Wei Wu, Qiang Wang, Fangyi Zhang, Junliang Xing, and Junjie Yan. Siamrpn++: Evolution of siamese vi- sual tracking with very deep networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4282–4291, 2019. 2, 6, 7

  22. [30]

    High performance visual tracking with siamese region pro- posal network

    Bo Li, Junjie Yan, Wei Wu, Zheng Zhu, and Xiaolin Hu. High performance visual tracking with siamese region pro- posal network. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8971–8980,

  23. [31]

    Visual object tracking for un- manned aerial vehicles: A benchmark and new motion mod- els

    Siyi Li and Dit-Yan Yeung. Visual object tracking for un- manned aerial vehicles: A benchmark and new motion mod- els. In AAAI, 2017. 6, 7

  24. [32]

    An informative track- ing benchmark

    Xin Li, Qiao Liu, Wenjie Pei, Qiuhong Shen, Yaowei Wang, Huchuan Lu, and Ming-Hsuan Yang. An informative track- ing benchmark. arXiv preprint arXiv:2112.06467, 2021. 6, 7

  25. [33]

    Revisiting batch normalization for practical do- main adaptation

    Yanghao Li, Naiyan Wang, Jianping Shi, Jiaying Liu, and Xiaodi Hou. Revisiting batch normalization for practical do- main adaptation. arXiv preprint arXiv:1603.04779, 2016. 2, 5, 7

  26. [34]

    Encoding color information for visual tracking: Algorithms and bench- mark

    Pengpeng Liang, Erik Blasch, and Haibin Ling. Encoding color information for visual tracking: Algorithms and bench- mark. IEEE transactions on image processing, 24(12):5630– 5644, 2015. 6, 7

  27. [35]

    Focal loss for dense object detection

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Doll´ar. Focal loss for dense object detection. In Pro- ceedings of the IEEE international conference on computer vision, pages 2980–2988, 2017. 5

  28. [36]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceeding...

  29. [37]

    Po- larized self-attention: Towards high-quality pixel-wise re- gression

    Huajun Liu, Fuqiang Liu, Xinyi Fan, and Dong Huang. Po- larized self-attention: Towards high-quality pixel-wise re- gression. arXiv preprint arXiv:2107.00782 , 2021. 2, 3, 5, 8

  30. [38]

    Transforming model prediction for tracking

    Christoph Mayer, Martin Danelljan, Goutam Bhat, Matthieu Paul, Danda Pani Paudel, Fisher Yu, and Luc Van Gool. Transforming model prediction for tracking. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8731–8740, 2022. 6

  31. [39]

    Separable self- attention for mobile vision transformers

    Sachin Mehta and Mohammad Rastegari. Separable self- attention for mobile vision transformers. arXiv preprint arXiv:2206.02680, 2022. 2, 3

  32. [40]

    Siamese network features for image matching

    Iaroslav Melekhov, Juho Kannala, and Esa Rahtu. Siamese network features for image matching. In 2016 23rd interna- tional conference on pattern recognition (ICPR), pages 378–

  33. [41]

    The norm must go on: Dynamic unsuper- vised domain adaptation by normalization

    M Jehanzeb Mirza, Jakub Micorek, Horst Possegger, and Horst Bischof. The norm must go on: Dynamic unsuper- vised domain adaptation by normalization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 14765–14775, 2022. 2, 5, 7

  34. [42]

    A benchmark and simulator for uav tracking

    Matthias Mueller, Neil Smith, and Bernard Ghanem. A benchmark and simulator for uav tracking. In Computer Vision–ECCV 2016: 14th European Conference, Amster- dam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14, pages 445–461. Springer, 2016. 6, 7

  35. [43]

    Trackingnet: A large-scale dataset and benchmark for object tracking in the wild

    Matthias Muller, Adel Bibi, Silvio Giancola, Salman Al- subaihi, and Bernard Ghanem. Trackingnet: A large-scale dataset and benchmark for object tracking in the wild. In Proceedings of the European conference on computer vision (ECCV), pages 300–317, 2018. 6, 7, 12, 13

  36. [44]

    Efficient test-time model adaptation without forgetting

    Shuaicheng Niu, Jiaxiang Wu, Yifan Zhang, Yaofo Chen, Shijian Zheng, Peilin Zhao, and Mingkui Tan. Efficient test-time model adaptation without forgetting. In Interna- tional conference on machine learning, pages 16888–16905. PMLR, 2022. 2, 5, 7

  37. [45]

    Avist: A benchmark for vi- sual object tracking in adverse visibility

    Mubashir Noman, Wafa Al Ghallabi, Daniya Najiha, Christoph Mayer, Akshay Dudhane, Martin Danelljan, Hisham Cholakkal, Salman Khan, Luc Van Gool, and Fahad Shahbaz Khan. Avist: A benchmark for vi- sual object tracking in adverse visibility. arXiv preprint arXiv:2208.06888, 2022...

  38. [46]

    Two at once: Enhancing learning and generalization capacities via ibn-net

    Xingang Pan, Ping Luo, Jianping Shi, and Xiaoou Tang. Two at once: Enhancing learning and generalization capacities via ibn-net. In Proceedings of the European Conference on Computer Vision (ECCV), pages 464–479, 2018. 2, 5, 7

  39. [47]

    Pytorch: An im- perative style, high-performance deep learning library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An im- perative style, high-performance deep learning library. Ad- vances in neural information processing systems ...

  40. [48]

    Generalized in- tersection over union: A metric and a loss for bounding box regression

    Hamid Rezatofighi, Nathan Tsoi, JunYoung Gwak, Amir Sadeghian, Ian Reid, and Silvio Savarese. Generalized in- tersection over union: A metric and a loss for bounding box regression. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 658–666,

  41. [49]

    Improving robustness against common corruptions by covariate shift adaptation

    Steffen Schneider, Evgenia Rusak, Luisa Eck, Oliver Bring- mann, Wieland Brendel, and Matthias Bethge. Improving robustness against common corruptions by covariate shift adaptation. Advances in neural information processing sys- tems, 33:11539–11551, 2020. 2, 5, 7

  42. [50]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia 10 Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. 3

  43. [51]

    Mobilevitv3: Mobile-friendly vision transformer with simple and effec- tive fusion of local, global and input features

    Shakti N Wadekar and Abhishek Chaurasia. Mobilevitv3: Mobile-friendly vision transformer with simple and effec- tive fusion of local, global and input features. arXiv preprint arXiv:2209.15159, 2022. 2, 3, 5

  44. [52]

    Tent: Fully test-time adaptation by entropy minimization

    Dequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno Ol- shausen, and Trevor Darrell. Tent: Fully test-time adaptation by entropy minimization. arXiv preprint arXiv:2006.10726,

  45. [53]

    Continual test-time domain adaptation

    Qin Wang, Olga Fink, Luc Van Gool, and Dengxin Dai. Continual test-time domain adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7201–7211, 2022. 2

  46. [54]

    Fast online object tracking and segmentation: A unifying approach

    Qiang Wang, Li Zhang, Luca Bertinetto, Weiming Hu, and Philip HS Torr. Fast online object tracking and segmentation: A unifying approach. In Proceedings of the IEEE/CVF con- ference on Computer Vision and Pattern Recognition, pages 1328–1338, 2019. 2, 6

  47. [55]

    Autoregressive visual tracking

    Xing Wei, Yifan Bai, Yongchao Zheng, Dahu Shi, and Yi- hong Gong. Autoregressive visual tracking. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 9697–9706, 2023. 1, 2

  48. [56]

    Cbam: Convolutional block attention module

    Sanghyun Woo, Jongchan Park, Joon-Young Lee, and In So Kweon. Cbam: Convolutional block attention module. In Proceedings of the European conference on computer vision (ECCV), pages 3–19, 2018. 2, 3

  49. [57]

    Fbnet: Hardware-aware efficient con- vnet design via differentiable neural architecture search

    Bichen Wu, Xiaoliang Dai, Peizhao Zhang, Yanghan Wang, Fei Sun, Yiming Wu, Yuandong Tian, Peter Vajda, Yangqing Jia, and Kurt Keutzer. Fbnet: Hardware-aware efficient con- vnet design via differentiable neural architecture search. In Proceedings of the IEEE/CVF conference on c...

  50. [58]

    Dropmae: Masked autoen- coders with spatial-attention dropout for tracking tasks

    Qiangqiang Wu, Tianyu Yang, Ziquan Liu, Baoyuan Wu, Ying Shan, and Antoni B Chan. Dropmae: Masked autoen- coders with spatial-attention dropout for tracking tasks. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 14561–14571, 2023. 1, 6, 7

  51. [59]

    Object track- ing benchmark

    Yi Wu, Jongwoo Lim, and Ming-Hsuan Yang. Object track- ing benchmark. IEEE Transactions on Pattern Analysis and Machine Intelligence, 37(9):1834–1848, 2015. 6, 7

  52. [60]

    Siamfc++: Towards robust and accurate visual tracking with target estimation guidelines

    Yinda Xu, Zeyu Wang, Zuoxin Li, Ye Yuan, and Gang Yu. Siamfc++: Towards robust and accurate visual tracking with target estimation guidelines. InProceedings of the AAAI con- ference on artificial intelligence , volume 34, pages 12549– 12556, 2020. 2

  53. [61]

    Learning spatio-temporal transformer for vi- sual tracking

    Bin Yan, Houwen Peng, Jianlong Fu, Dong Wang, and Huchuan Lu. Learning spatio-temporal transformer for vi- sual tracking. In Proceedings of the IEEE/CVF international conference on computer vision , pages 10448–10457, 2021. 1, 2, 6, 7

  54. [62]

    Lighttrack: Finding lightweight neural net- works for object tracking via one-shot architecture search

    Bin Yan, Houwen Peng, Kan Wu, Dong Wang, Jianlong Fu, and Huchuan Lu. Lighttrack: Finding lightweight neural net- works for object tracking via one-shot architecture search. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 15180–151...

  55. [63]

    Siamatt: Siamese attention network for visual tracking

    Kai Yang, Zhenyu He, Zikun Zhou, and Nana Fan. Siamatt: Siamese attention network for visual tracking. Knowledge- based systems, 203:106079, 2020. 2

  56. [64]

    Joint feature learning and relation modeling for tracking: A one-stream framework

    Botao Ye, Hong Chang, Bingpeng Ma, Shiguang Shan, and Xilin Chen. Joint feature learning and relation modeling for tracking: A one-stream framework. In European Conference on Computer Vision, pages 341–357. Springer, 2022. 6

  57. [65]

    Deformable siamese attention networks for visual ob- ject tracking

    Yuechen Yu, Yilei Xiong, Weilin Huang, and Matthew R Scott. Deformable siamese attention networks for visual ob- ject tracking. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6728–6737,

  58. [66]

    Ocean: Object-aware anchor-free tracking

    Zhipeng Zhang, Houwen Peng, Jianlong Fu, Bing Li, and Weiming Hu. Ocean: Object-aware anchor-free tracking. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXI 16, pages 771–787. Springer, 2020. 4, 6, 7

  59. [67]

    Distractor-aware siamese networks for visual object tracking

    Zheng Zhu, Qiang Wang, Bo Li, Wei Wu, Junjie Yan, and Weiming Hu. Distractor-aware siamese networks for visual object tracking. In Proceedings of the European conference on computer vision (ECCV), pages 101–117, 2018. 2 11 Algorithm 1 Sampling Strategy X ← rand(X n)|X ∈ X n ▷ ...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.