Pith. sign in

REVIEW 2 major objections 5 minor 60 references

Efficient Self-Supervised Video Hashing with Selective State Spaces

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read S5VH shows Mamba backbones can beat Transformers in self-supervised video hashing.

desk verdict First Mamba-based SSVH with solid experiments, but the hash-center ADMM update in Eq. (16) is the sphere projection, not the box projection, so the binary-center claim is not justified; needs a fix before the LCA contribution can be trusted. read the letter →

arxiv 2412.14518 v1 pith:W3ZT2HWC submitted 2024-12-19 cs.CV cs.IRcs.MM

Jinpeng Wang , Niu Lian , Jun Li , Yuting Wang , Yan Feng , Bin Chen , Yongbing Zhang , Shu-Tao Xia This is my paper · ORCID
classification cs.CVcs.IRcs.MM
keywords Mambastatespacemodelsself-supervisedvideohashinghashcenteralignmentretrievalHammingbidirectionalcontrastivelearning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to show that Mamba, a selective state-space model with linear-time sequence processing, can serve as the backbone for self-supervised video hashing, replacing the quadratic-cost Transformers that currently dominate the task. It introduces S5VH, whose encoder and decoder are built from bidirectional Mamba layers, and pairs it with a self-local-global learning scheme: masked-frame reconstruction gives a self signal, contrastive learning gives an inter-sample local signal, and a hash center alignment loss adds a corpus-level global signal. The global signal is obtained by clustering temporally averaged CNN features with k-means and then solving a constrained optimization that converts the cluster centroids into semantically consistent, well-separated binary hash centers. If the paper is right, state-space models offer both better retrieval accuracy and a scalable inference regime for video hashing, and hash center alignment is a reusable way to inject global semantics into self-supervised hash learning.

What carries the argument

The machinery has two parts. First, the bidirectional Mamba layer: each layer sums a forward and a backward Mamba block, where each block is a gated selective scan (linear projection, layer norm, 1D convolution, SiLU, and a selective state-space module with input-dependent $\Delta$, $B$, and $C$); this carries temporal modeling at linear complexity in frame number. Second, semantic hash center generation: k-means on temporally averaged video features yields $N_c$ centroids, and an $\ell^p$-box ADMM solver minimizes $\|\Phi\Phi^\top - K W\|_F^2 + \tfrac{1}{2}\sum_{i,j}\phi_i^\top\phi_j$ over binary $\Phi$, where $W$ is the cosine similarity matrix of centroids; the resulting hash centers are both consistent with feature-space semantics and mutually separated. The center alignment loss $\mathcal{L}_{CA} = -\log \frac{\exp(\phi_{c_i}^\top b_i/(K\tau))}{\sum_{c}\exp(\phi_c^\top b_i/(K\tau))}$ then pulls each video's hash code toward its assigned center during training.

What would settle it

Run S5VH on a video dataset whose category structure is not captured by the pretrained CNN features, and compare it with and without the center alignment loss; if the GmAP gap vanishes or reverses, the semantic meaningfulness of the k-means pseudo-labels is what carries the improvement. A cleaner control is to replace the real k-means centroids with random binary centers: if the retrieval gain survives, the global signal is doing the work, not the semantics.

Watch

Extended reading notes

Core claim

The central claim is that S5VH is the first Mamba-based self-supervised video hashing model and that it outperforms the state of the art on ActivityNet, FCVID, UCF101, and HMDB51 at 16-, 32-, and 64-bit code lengths, transfers better across datasets, and scales better at inference. The reported gains include a GmAP of 0.0378 in cross-dataset transfer from UCF101 to HMDB51, versus 0.0335 for the strongest baseline, and a fitted inference-time scaling law that is linear in frame count where the Transformer and MLP-Mixer baselines are quadratic. The paper also claims that the proposed center alignment loss, which aligns each video's hash code to the binary hash center of its k-means pseudo-class, contributes roughly six percent GmAP improvement and about 65 percent training-time acceleration, making the self-local-global paradigm the reason S5VH converges faster and better.

Load-bearing premise

The load-bearing premise is that k-means clusters of temporally averaged, pretrained CNN features correspond to semantic categories; if the clusters are noisy or semantically meaningless, the center alignment loss will pull hash codes toward arbitrary centroids and the reported gains from LCA would disappear.

Editorial extensions

If this is right

  • S5VH outperforms six baseline methods in mAP@N and precision-recall across ActivityNet, FCVID, UCF101, and HMDB51 at 16-, 32-, and 64-bit code lengths.
  • In cross-dataset transfer, training on UCF101 and testing on HMDB51 yields a GmAP of 0.0378 versus 0.0335 for the strongest baseline.
  • Inference time scales linearly with video length, with a fitted cost of roughly $1.5\times10^{-3}L$ milliseconds, whereas the Transformer and MLP-Mixer baselines fit quadratic scaling laws.
  • The center alignment loss contributes about six percent GmAP and roughly 65 percent faster convergence, so the global signal is a substantial efficiency lever during training.
  • Ablations show bidirectional Mamba beats forward-only and backward-only variants, and beats LSTM, RetNet, and RWKV when substituted into the same framework.

Reading between the lines

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

  • Hash center alignment is a plug-in training signal: it could be dropped into non-Mamba self-supervised hashers whose codes are compared by cosine or Hamming similarity, and the ablation pattern suggests it would help any such model.
  • The k-means pseudo-labels and ADMM hash centers are fixed before training, so a natural variant the paper does not test is re-estimating them periodically as features drift during training.
  • The linear-complexity inference result suggests state-space backbones are worth trying in other long-sequence retrieval tasks, such as near-duplicate video search or dense video captioning retrieval.
  • The gain from center alignment likely depends on how well the cluster count $N_c$ matches the true semantic structure; the paper fixes $N_c$ by dataset scale, so sensitivity to $N_c$ is a testable next step.
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

2 major / 5 minor

Summary. The paper presents S5VH, a self-supervised video hashing model built from bidirectional Mamba layers, with a self-local-global (SLG) learning paradigm that combines masked temporal reconstruction, contrastive learning, and a hash-center alignment loss. The hash centers are derived from k-means clustering of temporally averaged CNN features, followed by an ADMM optimization that aims to map cluster centroids to well-separated binary codes. Experiments on ActivityNet, FCVID, UCF101, and HMDB51 report improved retrieval accuracy over six baselines, cross-dataset transferability, and lower inference cost with linear scaling in sequence length. The paper includes ablations of the bidirectional design, loss terms, and SSM variants, and the code is publicly available.

Significance. If the empirical claims hold, S5VH is a meaningful step for efficient self-supervised video hashing, being the first Mamba-based model in this area. The linear-complexity inference and the global hash-center alignment signal are both practically motivated, and the paper provides a useful set of ablations. The public code and the ablation coverage are strengths. However, the central learning signal relies on an ADMM derivation that is currently incorrect as written, and the experimental section lacks measures of variance, so the significance can only be confirmed after a carefully revised version.

major comments (2)
  1. [Semantic Hash Center Generation, Eq. (16)] The update for the box-constraint auxiliary variable Ψ_b is written as the projection onto the Frobenius sphere of radius sqrt(NcK): Ψ_b^{k+1} = sqrt(NcK) * (Φ^{k+1} + Υ_b^k/μ_b) / ||Φ^{k+1} + Υ_b^k/μ_b||_F. This is the projection for the ℓ2-norm constraint S_p, not for the box constraint S_b ≡ [-1,+1]^{Nc×K}. The correct proximal update for S_b is elementwise clipping of Φ^{k+1} + Υ_b^k/μ_b to [-1,1]. As written, the ADMM never enforces the box constraint, so the output of the hash center generation is not guaranteed to be binary. Consequently, the simplification of the cosine similarity in Eq. (21) to φ_c^T b_i / (Kτ), which relies on both φ_c and b_i being ±1 vectors, is invalid unless the centers are actually binarized by some other mechanism. Please correct the update (assigning the sphere projection to the Ψ_p update), verify the actual outputs of the optimization, and report the fraction of ±1 entries or the constraint violation. The method is not reproducible as currently described.
  2. [Tables 1 and 2; Figure 3] All retrieval numbers are reported as single runs without standard deviations or statistical significance tests. The cross-dataset gain in Table 1 (e.g., 0.0285 vs 0.0378 on 16-bit codes) and the LCA ablation in Table 2 (e.g., 0.357 vs 0.286 at 16 bits on UCF101) could be affected by run-to-run variation. Since the central claim is that S5VH outperforms state-of-the-art methods, please provide mean ± std over at least three independent runs, or justify why a single run is representative. This is important for evaluating whether the reported improvements are reliable.
minor comments (5)
  1. [Eq. (25)] The metric is called 'geomean' (GmAP) but Eq. (25) computes the root-sum-square (Euclidean norm) of the mAP@N values, not the geometric mean. The geometric mean over the six N values would be the sixth root of the product of the mAP@N values. Please rename the metric or correct the formula, and adjust the captions and text of Figure 1 and Table 2 accordingly.
  2. [Implementation Details] The values of the loss weights α and β in Eq. (22) are not reported, although they are essential for reproducibility. Please list them, together with the early-stopping criteria and the number of runs.
  3. [References and figure citations] There are several presentation errors: the inference efficiency results are referred to as Figure 1(b) in the Experiments section, but Figure 1 shows inference efficiency in (a); the Model Analyses section refers to 'Figure 1(c)', which does not exist; the footnote defines the abbreviation as 'SSSSSVH' (an extra S); and 'stimulate tensors' should be 'simulate tensors'.
  4. [Eq. (20)] The contrastive loss in Eq. (20) is typeset in a way that is difficult to parse. Please rewrite it as the sum of two log-softmax terms, one treating the first view as anchor and the second treating the second view as anchor, to make the objective unambiguous.
  5. [Hyperparameter sensitivity] The number of semantic centers Nc and the mask ratio ρ are fixed per dataset, but no sensitivity analysis or justification is provided. A brief discussion of how these choices affect the hash center quality and retrieval performance would strengthen the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: hash centers come from external k-means features; retrieval is evaluated on ground-truth labels.

full rationale

The claimed derivation chain is self-contained. The pseudo-labels and hash centers are derived by k-means clustering of temporally averaged, pre-trained CNN features and by an ADMM optimization over those clusters; the center alignment loss then trains hash codes toward those centers. Evaluation is performed against ground-truth category labels, so the target result is never fed into the loss or the center-generation procedure. The Mamba encoder/decoder and the three losses are implemented and compared against rerun baselines rather than being assumed from citations. Self-citations, such as following ConMH for masked-frame reconstruction, are used only as a baseline and a standard design template and are not load-bearing for the central claims; the ablations show that the main gains come from the Mamba backbone and the center alignment loss, which are independently evaluated. The apparent inconsistency in Eq. (16), where the box-constraint update is written as a sphere projection, is a correctness/reproducibility concern rather than a circularity, because it does not reduce any prediction to the paper's own inputs.

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

The method rests on standard deep learning building blocks plus domain assumptions about the semantic quality of pretrained features and k-means clusters. No exotic entities are introduced. The main free parameters are loss balances and cluster counts, all chosen by hand.

free parameters (5)
  • alpha, beta loss weights = not reported
    Balance reconstruction, contrastive, and center-alignment losses in Equation (22); values not listed in the paper.
  • temperature tau = 0.5
    Used in contrastive and center alignment softmax losses (Equations 20 and 21).
  • mask ratio rho = 0.75 FCVID, 0.5 others
    Fraction of frames masked in the reconstruction loss (Equation 19); tuned per dataset.
  • number of semantic centers Nc = 450 FCVID, 100 others
    Number of k-means clusters and hash centers; chosen per dataset.
  • latent dimensions and layer counts = encoder 256, decoder 192; 6 encoder layers, 1 decoder layer
    Architecture hyperparameters set to keep model size comparable to baselines.
assumptions (5)
  • domain assumption Pre-trained ImageNet CNN features (ResNet-50, VGG-16) are semantically meaningful enough for clustering and reconstruction.
    The entire SLG paradigm starts from features extracted by frozen CNNs; if these features lack semantics, cluster pseudo-labels and hash centers are arbitrary.
  • domain assumption k-means on temporally averaged features compresses corpus-level semantic structure.
    Used to define pseudo-labels and hash centers before training; no evidence beyond the benchmarks.
  • domain assumption Bidirectional Mamba layers capture temporal dependencies at least as well as self-attention for video hashing.
    Motivates replacing Transformers; supported only by the experiments in this paper.
  • standard math The ℓp-box ADMM equivalence from Wu and Ghanem (2018) solves the binary hash center optimization.
    Used for Equations 11-18; assumed correct and convergent.
  • standard math Straight-through estimator can propagate gradients through the sign function.
    Used in Equation (10) for end-to-end hash learning.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Efficient Self-Supervised Video Hashing with Selective State Spaces." pith.science (2026). https://pith.science/paper/W3ZT2HWC

@misc{pith2026241214518,
  author       = {Pith},
  title        = {Pith review of: Efficient Self-Supervised Video Hashing with Selective State Spaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W3ZT2HWC}},
  note         = {Machine review of arXiv:2412.14518}
}
read the original abstract

Self-supervised video hashing (SSVH) is a practical task in video indexing and retrieval. Although Transformers are predominant in SSVH for their impressive temporal modeling capabilities, they often suffer from computational and memory inefficiencies. Drawing inspiration from Mamba, an advanced state-space model, we explore its potential in SSVH to achieve a better balance between efficacy and efficiency. We introduce S5VH, a Mamba-based video hashing model with an improved self-supervised learning paradigm. Specifically, we design bidirectional Mamba layers for both the encoder and decoder, which are effective and efficient in capturing temporal relationships thanks to the data-dependent selective scanning mechanism with linear complexity. In our learning strategy, we transform global semantics in the feature space into semantically consistent and discriminative hash centers, followed by a center alignment loss as a global learning signal. Our self-local-global (SLG) paradigm significantly improves learning efficiency, leading to faster and better convergence. Extensive experiments demonstrate S5VH's improvements over state-of-the-art methods, superior transferability, and scalable advantages in inference efficiency. Code is available at https://github.com/gimpong/AAAI25-S5VH.

Figures

Figures reproduced from arXiv: 2412.14518 by the authors.

Figure 1
Figure 1. Highlights: (a) Our S5VH based on Mamba ex￾hibits lower inference overheads on memory and computa￾tion. The efficiency advantages are scalable and more no￾table under larger frame numbers. (b) The introduced global learning signal in the hash space effectively enhances train￾ing efficiency, showing faster and better convergence. et al. 2018a; Li et al. 2019a) utilized nonlinear RNNs to pro￾cess frame features, which… view at source ↗
Figure 2
Figure 2. Overview of S5VH (best viewed in color). ( [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Retrieval PR curves of different models on the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (2 more)
Figure 3
Figure 3. Figure 3: Retrieval performance comparison by mAP@ [PITH_FULL_IMAGE:figures/full_fig_p006_3.png]
Figure 5
Figure 5. Figure 5: The t-SNE visualization of the learned hash codes [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 36 canonical work pages

  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]

    Bengio, Y.; L \'e onard, N.; and Courville, A. 2013. Estimating or propagating gradients through stochastic neurons for conditional computation. arXiv preprint arXiv:1308.3432

    arXiv 2013

  4. [4]

    Caba Heilbron, F.; Escorcia, V.; Ghanem, B.; and Carlos Niebles, J. 2015. Activitynet: A large-scale video benchmark for human activity understanding. In Proceedings of the ieee conference on computer vision and pattern recognition, 961--970

    work page 2015

  5. [5]

    Duan, J.; Hao, Y.; Zhu, B.; Cheng, L.; Zhou, P.; and Wang, X. 2024. Efficient Unsupervised Video Hashing with Contextual Modeling and Structural Controlling. IEEE Transactions on Multimedia

    work page 2024

  6. [6]

    Y.; Dao, T.; Saab, K

    Fu, D. Y.; Dao, T.; Saab, K. K.; Thomas, A. W.; Rudra, A.; and R \'e , C. 2022. Hungry hungry hippos: Towards language modeling with state space models. arXiv preprint arXiv:2212.14052

    arXiv 2022

  7. [7]

    Gao, K.; Bai, J.; Chen, B.; Wu, D.; and Xia, S.-T. 2023. Backdoor Attack on Hash-based Image Retrieval via Clean-label Data Poisoning. In The 34th British Machine Vision Conference, 172--173

    work page 2023

  8. [8]

    Gong, Y.; Lazebnik, S.; Gordo, A.; and Perronnin, F. 2012. Iterative quantization: A procrustean approach to learning binary codes for large-scale image retrieval. IEEE transactions on pattern analysis and machine intelligence, 35(12): 2916--2929

    2012

Show all 60 references
  1. [9]

    Gu, A.; and Dao, T. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752

    2023 arXiv

  2. [10]

    Gu, A.; Dao, T.; Ermon, S.; Rudra, A.; and R \'e , C. 2020. Hippo: Recurrent memory with optimal polynomial projections. Advances in Neural Information Processing Systems, 33

    2020

  3. [11]

    Gu, A.; Goel, K.; Gupta, A.; and R \'e , C. 2022. On the parameterization and initialization of diagonal state space models. Advances in Neural Information Processing Systems, 35: 35971--35983

    2022

  4. [12]

    Gu, A.; Goel, K.; and R \'e , C. 2021. Efficiently Modeling Long Sequences with Structured State Spaces. In International Conference on Learning Representations

    2021

  5. [13]

    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. Advances in neural information processing systems, 34: 572--585

    2021

  6. [14]

    Gupta, A.; Gu, A.; and Berant, J. 2022. Diagonal state spaces are as effective as structured state spaces. Advances in Neural Information Processing Systems, 35: 22982--22994

    2022

  7. [15]

    Hao, Y.; Duan, J.; Zhang, H.; Zhu, B.; Zhou, P.; and He, X. 2022. Unsupervised video hashing with multi-granularity contextualization and multi-structure preservation. In Proceedings of the 30th ACM International Conference on Multimedia, 3754--3763

    2022

  8. [16]

    Y.; Jiang, J.; Hong, R.; and Wang, M

    Hao, Y.; Mu, T.; Goulermas, J. Y.; Jiang, J.; Hong, R.; and Wang, M. 2017. Unsupervised t-distributed video hashing and its deep hashing extension. IEEE Transactions on Image Processing, 26(11): 5531--5544

    2017

  9. [17]

    He, K.; Fan, H.; Wu, Y.; Xie, S.; and Girshick, R. 2020. Momentum contrast for unsupervised visual representation learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 9729--9738

    2020

  10. [18]

    He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, 770--778

    2016

  11. [19]

    Hochreiter, S.; and Schmidhuber, J. 1997. Long Short-Term Memory. Neural Computation, 9(8): 1735--1780

    1997

  12. [20]

    Jiang, Y.-G.; Wu, Z.; Wang, J.; Xue, X.; and Chang, S.-F. 2017. Exploiting feature and class relationships in video categorization with regularized deep neural networks. IEEE transactions on pattern analysis and machine intelligence, 40(2): 352--364

    2017

  13. [21]

    Kalman, R. E. 1960. A New Approach to Linear Filtering and Prediction Problems . Journal of Basic Engineering, 82(1): 35--45

    1960

  14. [22]

    A.; and Serre, T

    Kuehne, H.; Jhuang, H.; Garrote, E.; Poggio, T. A.; and Serre, T. 2011. HMDB: A large video database for human motion recognition. In Metaxas, D. N.; Quan, L.; Sanfeliu, A.; and Gool, L. V., eds., IEEE International Conference on Computer Vision, ICCV 2011, Barcelona, Spain, N...

    2011

  15. [23]

    Li, C.; Yang, Y.; Cao, J.; and Huang, Z. 2017. Jointly modeling static visual appearance and temporal pattern for unsupervised video hashing. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management, 9--17

    2017

  16. [24]

    Li, P.; Xie, H.; Ge, J.; Zhang, L.; Min, S.; and Zhang, Y. 2022. Dual-stream knowledge-preserving hashing for unsupervised video retrieval. In European Conference on Computer Vision, 181--197. Springer

    2022

  17. [25]

    Li, Q.; Tian, X.; and Ng, W. W. 2024. Self-supervised Temporal Sensitive Hashing for Video Retrieval. IEEE Transactions on Multimedia

    2024

  18. [26]

    Li, S.; Chen, Z.; Li, X.; Lu, J.; and Zhou, J. 2019 a . Unsupervised variational video hashing with 1D-CNN-LSTM networks. IEEE Transactions on Multimedia, 22(6): 1542--1554

    2019

  19. [27]

    Li, S.; Chen, Z.; Lu, J.; Li, X.; and Zhou, J. 2019 b . Neighborhood preserving hashing for scalable video retrieval. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 8212--8221

    2019

  20. [28]

    Li, S.; Li, X.; Lu, J.; and Zhou, J. 2021. Self-supervised video hashing via bidirectional transformers. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 13549--13558

    2021

  21. [29]

    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

  22. [30]

    Peng, B.; Alcaide, E.; Anthony, Q.; Albalak, A.; Arcadinho, S.; Biderman, S.; Cao, H.; Cheng, X.; Chung, M.; Derczynski, L.; et al. 2023. RWKV: Reinventing RNNs for the Transformer Era. In Findings of the Association for Computational Linguistics: EMNLP 2023, 14048--14077

    2023

  23. [31]

    Qiao, Y.; Yu, Z.; Guo, L.; Chen, S.; Zhao, Z.; Sun, M.; Wu, Q.; and Liu, J. 2024. Vl-mamba: Exploring state space models for multimodal learning. arXiv preprint arXiv:2403.13600

    2024 arXiv

  24. [32]

    Ramachandran, P.; Zoph, B.; and Le, Q. V. 2017. Searching for activation functions. arXiv preprint arXiv:1710.05941

    2017 arXiv

  25. [33]

    Shen, X.; Zhou, Y.; Yuan, Y.-H.; Yang, X.; Lan, L.; and Zheng, Y. 2023. Contrastive Transformer Hashing for Compact Video Representation. IEEE Transactions on Image Processing

    2023

  26. [34]

    Simonyan, K.; and Zisserman, A. 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556

    2014 arXiv

  27. [35]

    T.; Warrington, A.; and Linderman, S

    Smith, J. T.; Warrington, A.; and Linderman, S. W. 2022. Simplified state space layers for sequence modeling. arXiv preprint arXiv:2208.04933

    2022 arXiv

  28. [36]

    T.; and Hong, R

    Song, J.; Yang, Y.; Huang, Z.; Shen, H. T.; and Hong, R. 2011. Multiple feature hashing for real-time large scale near-duplicate video retrieval. In Proceedings of the 19th ACM international conference on Multimedia, 423--432

    2011

  29. [37]

    Song, J.; Zhang, H.; Li, X.; Gao, L.; Wang, M.; and Hong, R. 2018 a . Self-supervised video hashing with hierarchical binary auto-encoder. IEEE Transactions on Image Processing, 27(7): 3210--3221

    2018

  30. [38]

    Song, J.; Zhang, H.; Li, X.; Gao, L.; Wang, M.; and Hong, R. 2018 b . Self-supervised video hashing with hierarchical binary auto-encoder. IEEE Transactions on Image Processing, 27(7): 3210--3221

    2018

  31. [39]

    R.; and Shah, M

    Soomro, K.; Zamir, A. R.; and Shah, M. 2012. UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild. CoRR, abs/1212.0402

    2012 arXiv

  32. [40]

    Sun, Y.; Dong, L.; Huang, S.; Ma, S.; Xia, Y.; Xue, J.; Wang, J.; and Wei, F. 2023 a . Retentive network: A successor to transformer for large language models. arXiv preprint arXiv:2307.08621

    2023 arXiv

  33. [41]

    Sun, Y.; Qin, Y.; Peng, D.; Ren, Z.; Yang, C.; and Hu, P. 2024. Dual Self-Paced Hashing for Image Retrieval. IEEE Transactions on Multimedia, 26: 9619--9629

    2024

  34. [42]

    Sun, Y.; Ren, Z.; Hu, P.; Peng, D.; and Wang, X. 2023 b . Hierarchical consensus hashing for cross-modal retrieval. IEEE Transactions on Multimedia, 26: 824--836

    2023

  35. [43]

    O.; Houlsby, N.; Kolesnikov, A.; Beyer, L.; Zhai, X.; Unterthiner, T.; Yung, J.; Steiner, A.; Keysers, D.; Uszkoreit, J.; et al

    Tolstikhin, I. O.; Houlsby, N.; Kolesnikov, A.; Beyer, L.; Zhai, X.; Unterthiner, T.; Yung, J.; Steiner, A.; Keysers, D.; Uszkoreit, J.; et al. 2021. Mlp-mixer: An all-mlp architecture for vision. Advances in neural information processing systems, 34: 24261--24272

    2021

  36. [44]

    van der Maaten, L.; and Hinton, G. E. 2008. Visualizing Data using t-SNE. Journal of Machine Learning Research, 9: 2579--2605

    2008

  37. [45]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. Advances in neural information processing systems, 30

    2017

  38. [46]

    Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Li \`o , P.; and Bengio, Y. 2018. Graph Attention Networks. In International Conference on Learning Representations

    2018

  39. [47]

    Wang, C.; Tsepa, O.; Ma, J.; and Wang, B. 2024 a . Graph-mamba: Towards long-range graph sequence modeling with selective state spaces. arXiv preprint arXiv:2402.00789

    2024 arXiv

  40. [48]

    Wang, J.; Zeng, Z.; Chen, B.; Wang, Y.; Liao, D.; Li, G.; Wang, Y.; and Xia, S.-T. 2024 b . Hugs Bring Double Benefits: Unsupervised Cross-Modal Hashing with Multi-granularity Aligned Transformers. International Journal of Computer Vision, 1--33

    2024

  41. [49]

    Wang, Y.; Wang, J.; Chen, B.; Zeng, Z.; and Xia, S.-T. 2023 a . Contrastive masked autoencoders for self-supervised video hashing. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37(3), 2733--2741

    2023

  42. [50]

    Wang, Y.; Zhou, M.; Sun, Y.; and Qian, X. 2023 b . Uncertainty-aware unsupervised video hashing. In International Conference on Artificial Intelligence and Statistics, 6722--6740

    2023

  43. [51]

    Wei, R.; Liu, Y.; Song, J.; Cui, H.; Xie, Y.; and Zhou, K. 2023. CHAIN: Exploring Global-Local Spatio-Temporal Information for Improved Self-Supervised Video Hashing. In Proceedings of the 31st ACM International Conference on Multimedia, 1677--1688

    2023

  44. [52]

    Weiss, Y.; Torralba, A.; and Fergus, R. 2008. Spectral hashing. Advances in neural information processing systems, 21

    2008

  45. [53]

    Wu, B.; and Ghanem, B. 2018. _p -Box ADMM: A Versatile Framework for Integer Programming. IEEE transactions on pattern analysis and machine intelligence, 41(7): 1695--1708

    2018

  46. [54]

    Wu, G.; Liu, L.; Guo, Y.; Ding, G.; Han, J.; Shen, J.; and Shao, L. 2017. Unsupervised deep video hashing with balanced rotation. In Proceedings of the 26th International Joint Conference on Artificial Intelligence, 3076--3082

    2017

  47. [55]

    Ye, G.; Liu, D.; Wang, J.; and Chang, S.-F. 2013. Large-scale video hashing via structure learning. In Proceedings of the IEEE international conference on computer vision, 2272--2279

    2013

  48. [56]

    Zeng, Z.; Wang, J.; Chen, B.; Wang, Y.; and Xia, S.-T. 2022. Motion-Aware Graph Reasoning Hashing for Self-supervised Video Retrieval. In 33rd British Machine Vision Conference, 82

    2022

  49. [57]

    Zhang, H.; Wang, M.; Hong, R.; and Chua, T.-S. 2016. Play and rewind: Optimizing binary representations of videos by self-supervised temporal hashing. In Proceedings of the 24th ACM international conference on Multimedia, 781--790

    2016

  50. [58]

    Zhao, H.; Zhang, M.; Zhao, W.; Ding, P.; Huang, S.; and Wang, D. 2024. Cobra: Extending mamba to multi-modal large language model for efficient inference. arXiv preprint arXiv:2403.14520

    2024 arXiv

  51. [59]

    Zhou, Y.; Sun, Z.; Liu, R.; Chen, Y.; and Zhang, D. 2024. AVHash: Joint Audio-Visual Hashing for Video Retrieval. In Proceedings of the 32nd ACM international conference on Multimedia

    2024

  52. [60]

    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.