Pith. sign in

REVIEW 3 major objections 6 minor 65 references

MambaHash: Visual State Space Deep Hashing Model for Large-Scale Image Retrieval

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A state-space deep hashing model, MambaHash, claims the highest reported mean average precision on three image retrieval benchmarks across 16, 32, 48, and 64-bit hash codes.

desk verdict First Mamba-based deep hashing model with a coherent architecture and real efficiency gains, but the 'superior performance' claim rests on single-run margins as small as 0.0001 against the authors' own HybridHash baseline. read the letter →

arxiv 2506.16353 v1 pith:VPXWBGA5 submitted 2025-06-19 cs.CV

Chao He , Hongxi Wei This is my paper
classification cs.CV
keywords deephashingimageretrievalMambavisualstatespacemodelchannelinteractionattentionadaptivefeatureenhancementbinaryhashcodeslarge-scale
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

Large-scale image retrieval needs binary hash codes that are compact enough for fast search and accurate enough to return relevant images. This paper argues that Mamba, a state-space architecture with linear-time scanning, can power a deep hashing model if the network is structured to keep both local and global information. To support that claim, the authors build MambaHash, whose backbone splits channels into four groups, scans each group in a different direction, and adds a channel-interaction attention module and an adaptive feature-enhancement module. Trained with pairwise likelihood and quantization losses, MambaHash reports mean average precision (MAP) that is the highest among the compared methods on CIFAR-10 and NUS-WIDE at all four code lengths and on IMAGENET at 16, 32, and 64 bits, with the 48-bit IMAGENET result within 0.0007 of the best baseline. If the reported margins survive repeated runs, MambaHash would show that linear-complexity state-space models can replace quadratic self-attention in retrieval pipelines.

What carries the argument

The load-bearing mechanism is the grouped multi-directional Mamba scan. The backbone divides its channels into four groups and runs a VSSS block (the visual selective-scan building block) on each group in a distinct direction — left-to-right, right-to-left, top-to-bottom, bottom-to-top — then concatenates the results, giving global context with linear complexity in sequence length. The channel interaction attention module (CIAM) repairs the cross-channel communication that grouping cuts off: global average pooling feeds one $k\times k$ 1D convolution for local channel context and one linear layer for global context, and their sum is passed through a sigmoid to score channels. The adaptive feature enhancement module (AFEM) expands the channel dimension by the ratio $\lambda = 2^{\mu K + b}$, with $\mu=1/16$ and $b=0$, applies $1\times1$, $3\times3$, and $5\times5$ depthwise convolutions to capture multi-scale detail, and restores the original width. A hash layer with tanh outputs binary codes, trained by pairwise negative log-likelihood plus a quantization error term.

What would settle it

Re-run MambaHash and HybridHash on NUS-WIDE 16-bit and IMAGENET 16-bit and 48-bit with identical splits and at least five random seeds; if the MAP gaps (0.0001, 0.0105, and -0.0007, respectively) fall within one standard deviation, the claimed superiority over the strongest baseline is not established.

Watch

Extended reading notes

Core claim

The paper's central claim is that Mamba, despite being designed for long-sequence autoregressive tasks, is suitable for image retrieval when the network is built around grouped, multi-directional selective scans. MambaHash is that construction: a stage-wise backbone that splits channels into four groups, scans them left-to-right, right-to-left, top-to-bottom, and bottom-to-top, then reconnects the groups with a channel-interaction attention module and diversifies features with an adaptive enhancement module. Trained on pairwise images with a negative log-likelihood objective plus quantization error, it produces binary hash codes that the authors report as superior to the compared deep supervised hashing methods on CIFAR-10, NUS-WIDE, and IMAGENET. The ablations support the design logic: removing either module lowers MAP everywhere, and the Mamba-only variant remains competitive, which the authors take as evidence that Mamba itself can carry an image retrieval pipeline.

Load-bearing premise

The load-bearing premise is that the reported MAP differences over the strongest baseline, several of which are only 0.0001 to 0.005, reflect real model gains rather than run-to-run variation, given that all baseline numbers were copied from an earlier paper and no repeated trials or error bars are reported.

Editorial extensions

If this is right

  • At 48 bits, MambaHash reports 38.99M parameters and 7.53G FLOPs, against 55.57M parameters and 14.17G FLOPs for the hybrid Transformer-CNN baseline, so the linear-complexity backbone is the cheaper option in the comparison.
  • The ablation results show both added modules pull weight: removing CIAM or AFEM lowers MAP on every dataset and bit length, while the plain Mamba variant still holds its own.
  • The same architecture ranks first or second on every benchmark configuration in Table 1, which suggests the design transfers across single-label and multi-label retrieval.
  • The hash-length-dependent expansion rule $\lambda = 2^{\mu K + b}$ offers a fixed recipe for scaling feature capacity with code length, avoiding per-bit manual tuning.

Reading between the lines

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

  • The authors take all baseline numbers from their own earlier HybridHash paper rather than re-running them; if the two models are trained in the same code with multiple seeds, the margins on NUS-WIDE (as small as 0.0001 MAP) may not hold, making the efficiency advantage the more durable result.
  • The reported FLOP and parameter counts compare backbone architectures, not end-to-end retrieval latency; a straightforward extension is to measure query time and memory on a million-scale database, where Mamba's linear complexity should widen the gap.
  • Because $\mu$ and $b$ in the adaptive expansion rule were fixed once and applied across all experiments, tuning them per dataset is an obvious extension that could either enlarge or shrink the observed gains.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes MambaHash, a deep hashing model built on a visual state space backbone. The backbone uses grouped Mamba scanning in four directions, a channel interaction attention module (CIAM) for cross-channel communication, and an adaptive feature enhancement module (AFEM) that scales channel expansion with hash code length. The model is trained with a pairwise negative log-likelihood loss plus a quantization loss. Experiments on CIFAR-10, NUS-WIDE, and IMAGENET report MAP at 16, 32, 48, and 64 bits, and the authors claim that MambaHash achieves superior performance and efficiency compared with state-of-the-art deep hashing methods.

Significance. If the empirical claims hold, MambaHash would be the first Mamba-based deep hashing model and would show that selective state space backbones can be competitive for large-scale image retrieval while maintaining linear complexity. The architecture is internally coherent, the ablations in Table 2 demonstrate that CIAM and AFEM each contribute, and the authors release source code. However, the central 'superior performance' claim currently rests on single-run comparisons with very small margins against the strongest baseline, and one reported cell is actually below the baseline; the significance is therefore conditional on a substantially strengthened evaluation protocol.

major comments (3)
  1. [§4.3, Table 1] The claim that MambaHash has 'superior performance' over state-of-the-art deep hashing methods is not supported by the reported NUS-WIDE margins against HybridHash: 0.0001, 0.0005, 0.0007, and 0.0014 at 16, 32, 48, and 64 bits. The paper reports a single run, provides no variance or seed information, and states in §4.3 that all baseline results are 'derived from [20]', the authors' own prior paper. Differences of this magnitude are within typical run-to-run noise and also inherit any protocol differences in [20]. Moreover, the IMAGENET 48-bit entry is lower than HybridHash (0.9087 vs. 0.9094), directly contradicting the assertion that MambaHash consistently outperforms. Please report multiple seeds with means and standard deviations, perform significance tests where appropriate, and ideally re-run the strongest baselines under the same protocol.
  2. [§3.4, Eq. (16)] The adaptive channel enhancement rule λ = 2^(μK+b), with μ=1/16 and b=0, is introduced without sensitivity analysis. Since AFEM is presented as a main contribution, the adaptive rule should be validated against fixed ratios or alternative mappings (e.g., a linear mapping) to show that the gain comes from adaptivity rather than simply from increased channel capacity. Please add ablations over μ, b, and the CIAM kernel size k.
  3. [§4.2, Eq. (18)] The hyperparameter written as 'a' in §4.2 is never defined; presumably it is η in Eq. (18), the weight of the quantization loss. No cross-validation range or sensitivity analysis is reported for this hyperparameter, which is necessary to reproduce the trade-off between the pairwise NLL loss and the quantization loss. Please specify the exact hyperparameter and its tuning procedure.
minor comments (6)
  1. [§3.5] There is a typo in the text: 'Sicne' should be 'Since'.
  2. [§3.3.3, Eq. (11)] The activation function is written as 'Sigmod' and should be 'Sigmoid'.
  3. [Table 3 caption] The caption contains 'HybirdHash', which should be 'HybridHash'.
  4. [Table 1] The HybridHash row for IMAGENET is typeset without spaces: '0.88860.90940.9110' should read '0.8886 0.9094 0.9110'.
  5. [§4.4] The hidden feature dimensions are listed as 64, 128, 348, and 512; 348 is unusual and is likely a typo for 384 or 256. Please clarify.
  6. [Abstract] The phrase 'well efficiency and superior performance' is ungrammatical; consider 'good efficiency and superior performance'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the model is an architectural proposal evaluated by measured MAP, with no predicted quantity defined from a fitted parameter.

full rationale

The paper's central claim is empirical: MambaHash achieves competitive or superior MAP on CIFAR-10, NUS-WIDE, and IMAGENET. The derivation chain contains no step in which an output quantity is defined in terms of the very result it is supposed to predict. The AFEM channel-enhancement rule, λ=2^(μK+b) with μ=1/16 and b=0 (Eq. 16), is an explicitly stated design heuristic; it is not fitted to the reported MAP values, and the reported MAP is a measured outcome of training, not a consequence of this formula. The objective functions in Eqs. 17-18 are standard pairwise negative log-likelihood and quantization losses; they guide optimization but do not encode the final MAP numbers. The only notable self-citation is [20], the authors' prior HybridHash paper, from which all baseline results are taken (Section 4.3: 'all the results of non-deep-learning hashing methods and deep-learning hashing methods are derived from [20]'). This weakens the independence of the comparison, but it is not circular: MambaHash's MAP values are independently measured outputs of a trained network, not re-expressions of those baseline numbers or of any fitted hyperparameter. No equation in the paper reduces to its own input, and no claimed prediction is forced by construction. Therefore no circular step is exhibited.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper's performance claims rest mainly on standard SSM math and the pairwise hashing objective, both from prior work, plus four free hyperparameters guessed or cross-validated. The only ad hoc invention is the exponential channel enhancement rule, which is not derived. No new physical or domain entities are introduced.

free parameters (4)
  • eta (quantization loss weight) = 0.01 on CIFAR-10, 0.05 on NUS-WIDE and IMAGENET
    Cross-validated in Section 4.2; the paper says 'We obtained the hyperparameter a of MambaHash by cross-validation to be 0.01, 0.05, and 0.05 on the CIFAR-10, NUS-WIDE and IMAGENET datasets, respectively.' This per-dataset fitted value controls the objective in Eq. 18.
  • mu and b in channel enhancement ratio = mu = 1/16, b = 0
    Chosen by hand in Section 3.4; lambda = 2^(mu*K+b). No tuning or derivation is reported, yet this choice controls channel expansion in AFEM and therefore model capacity and results.
  • CIAM kernel size k = 3, 3, 5, 5 for 16/32/48/64 bits
    Set in Section 4.4 'with reference to the method in [47]', a heuristic from ECA-Net, not learned or justified for this task.
  • learning rate = within [1e-5, 2e-5], exact value not stated
    Tuned in Section 4.2; the exact value for each dataset is not reported, leaving a free experimental parameter.
assumptions (5)
  • standard math Zero-Order Hold discretization of continuous SSMs (Eq. 2-3) is a valid way to integrate Mamba layers into the network.
    Invoked in Section 3.1 Preliminaries, borrowed from S4/Mamba [15,16]; this is accepted background math.
  • domain assumption The pairwise negative log-likelihood loss plus quantization loss (Eqs. 17-18) is an effective objective for learning hash codes.
    Used from DPSH [31] and DHN [60]; the paper provides no comparison to other hashing losses, so the objective is assumed to be suitable.
  • domain assumption Mamba's selective scan can serve as a vision backbone for image retrieval when combined with grouped scanning.
    This is the central hypothesis the paper tests (Section 1: 'we argue that Mamba is still available for image retrieval tasks'); the performance results are the evidence, but the suitability is assumed a priori in the architecture design.
  • ad hoc to paper The exponential mapping lambda = 2^(mu*K+b) between hash code length and channel enhancement ratio is sensible and generalizes across datasets.
    Introduced in Section 3.4 with no derivation or ablation over mu and b; if this relationship is wrong, the AFEM module would not provide its claimed benefit, though the ablation shows it helps at the chosen setting.
  • domain assumption Pretrained Mamba backbone weights, if used, transfer to hashing.
    The paper does not state whether the backbone is pretrained on ImageNet or trained from scratch; if pretrained weights are used, transferability is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MambaHash: Visual State Space Deep Hashing Model for Large-Scale Image Retrieval." pith.science (2026). https://pith.science/paper/VPXWBGA5

@misc{pith2026250616353,
  author       = {Pith},
  title        = {Pith review of: MambaHash: Visual State Space Deep Hashing Model for Large-Scale Image Retrieval},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VPXWBGA5}},
  note         = {Machine review of arXiv:2506.16353}
}
read the original abstract

Deep image hashing aims to enable effective large-scale image retrieval by mapping the input images into simple binary hash codes through deep neural networks. More recently, Vision Mamba with linear time complexity has attracted extensive attention from researchers by achieving outstanding performance on various computer tasks. Nevertheless, the suitability of Mamba for large-scale image retrieval tasks still needs to be explored. Towards this end, we propose a visual state space hashing model, called MambaHash. Concretely, we propose a backbone network with stage-wise architecture, in which grouped Mamba operation is introduced to model local and global information by utilizing Mamba to perform multi-directional scanning along different groups of the channel. Subsequently, the proposed channel interaction attention module is used to enhance information communication across channels. Finally, we meticulously design an adaptive feature enhancement module to increase feature diversity and enhance the visual representation capability of the model. We have conducted comprehensive experiments on three widely used datasets: CIFAR-10, NUS-WIDE and IMAGENET. The experimental results demonstrate that compared with the state-of-the-art deep hashing methods, our proposed MambaHash has well efficiency and superior performance to effectively accomplish large-scale image retrieval tasks. Source code is available https://github.com/shuaichaochao/MambaHash.git

Figures

Figures reproduced from arXiv: 2506.16353 by the authors.

Figure 1
Figure 1. The direction of scanning along the four groups of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The detailed architecture of the proposed MambaHash. MambaHash accepts pairwise images as input, and adopts a [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

65 extracted references · 43 canonical work pages

  1. [20]

    Chao He and Hongxi Wei. 2024. HybridHash: Hybrid Convolutional and Self- Attention Deep Hashing for Image Retrieval. InProceedings of the 2024 Interna- tional Conference on Multimedia Retrieval. 824–832

    work page 2024

  2. [1]

    Syed Talal Wasim Abdelrahman Shaker, Salman Khan, Juergen Gall, and Fa- had Shahbaz Khan. 2024. Groupmamba: Parameter-efficient and accurate group visual state space model.arXiv preprint arXiv:2407.137722 (2024)

    arXiv 2024

  3. [2]

    Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E Hinton. 2016. Layer normaliza- tion.arXiv preprint arXiv:1607.06450(2016)

    arXiv 2016

  4. [3]

    Ali Behrouz, Michele Santacatterina, and Ramin Zabih. 2024. Mambamixer: Efficient selective state space models with dual token and channel selection. arXiv preprint arXiv:2403.19888(2024)

    arXiv 2024

  5. [4]

    Joel Brogan, Aparna Bharati, Daniel Moreira, Anderson Rocha, Kevin W Bowyer, Patrick J Flynn, and Walter J Scheirer. 2021. Fast local spatial verification for feature-agnostic large-scale image retrieval.IEEE Transactions on image processing 30 (2021), 6892–6905

    work page 2021

  6. [5]

    Yue Cao, Mingsheng Long, Bin Liu, and Jianmin Wang. 2018. Deep cauchy hashing for hamming space retrieval. InProceedings of the IEEE conference on computer vision and pattern recognition. 1229–1237

    work page 2018

  7. [6]

    Zhangjie Cao, Mingsheng Long, Jianmin Wang, and Philip S Yu. 2017. Hashnet: Deep learning to hash by continuation. InProceedings of the IEEE international conference on computer vision. 5608–5617

    2017

  8. [7]

    Moses S Charikar. 2002. Similarity estimation techniques from rounding algo- rithms. InProceedings of the thiry-fourth annual ACM symposium on Theory of computing. 380–388

    2002

Show all 65 references
  1. [8]

    Yongbiao Chen, Sheng Zhang, Fangxin Liu, Zhigang Chang, Mang Ye, and Zheng- wei Qi. 2022. Transhash: Transformer-based hamming hashing for efficient image retrieval. InProceedings of the International Conference on Multimedia Retrieval. 127–136

    2022

  2. [9]

    Xiangxiang Chu, Zhi Tian, Yuqing Wang, Bo Zhang, Haibing Ren, Xiaolin Wei, Huaxia Xia, and Chunhua Shen. 2021. Twins: Revisiting the design of spatial attention in vision transformers.Advances in neural information processing systems34 (2021), 9355–9366

    2021

  3. [10]

    Tat-Seng Chua, Jinhui Tang, Richang Hong, Haojie Li, Zhiping Luo, and Yantao Zheng. 2009. Nus-wide: a real-world web image database from national university of singapore. InProceedings of the ACM international conference on image and video retrieval. 1–9

    2009

  4. [11]

    Hui Cui, Lei Zhu, Jingjing Li, Yang Yang, and Liqiang Nie. 2019. Scalable deep hashing for large-scale social image retrieval.IEEE Transactions on image pro- cessing29 (2019), 1271–1284

    2019

  5. [12]

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xi- aohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. 2021. An image is worth 16x16 words: Transformers for image recognition at scale. InProceeding...

    2021

  6. [13]

    Lixin Fan, Kam Woh Ng, Ce Ju, Tianyu Zhang, and Chee Seng Chan. 2020. Deep Polarized Network for Supervised Learning of Accurate Binary Hashing Codes. In Proceedings of the International Joint Conference on Artificial Intelligence. 825–831

    2020

  7. [14]

    Yunchao Gong, Svetlana Lazebnik, Albert Gordo, and Florent Perronnin. 2012. Iterative quantization: A procrustean approach to learning binary codes for large-scale image retrieval.IEEE Transactions on Pattern Analysis and Machine Intelligence35, 12 (2012), 2916–2929

    2012

  8. [15]

    Albert Gu and Tri Dao. 2023. Mamba: Linear-time sequence modeling with selective state spaces.arXiv preprint arXiv:2312.00752(2023)

    2023 arXiv

  9. [16]

    Albert Gu, Karan Goel, and Christopher Ré. 2021. Efficiently modeling long sequences with structured state spaces.arXiv preprint arXiv:2111.00396(2021)

    2021 arXiv

  10. [17]

    Albert Gu, Isys Johnson, Karan Goel, Khaled Saab, Tri Dao, Atri Rudra, and Christopher Ré. 2021. Combining recurrent, convolutional, and continuous-time models with linear state space layers.Advances in neural information processing systems34 (2021), 572–585

    2021

  11. [18]

    Jianyuan Guo, Kai Han, Han Wu, Yehui Tang, Xinghao Chen, Yunhe Wang, and Chang Xu. 2022. Cmt: Convolutional neural networks meet vision transform- ers. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 12175–12185

    2022

  12. [19]

    Yuchen Guo, Guiguang Ding, Li Liu, Jungong Han, and Ling Shao. 2017. Learning to hash with optimized anchor embedding for scalable retrieval.IEEE Transactions on image processing26 (2017), 1344–1354

    2017

  13. [21]

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep residual learning for image recognition. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 770–778

    2016

  14. [22]

    Tong He, Zhi Zhang, Hang Zhang, Zhongyue Zhang, Junyuan Xie, and Mu Li

  15. [23]

    Xuanhua He, Ke Cao, Jie Zhang, Keyu Yan, Yingying Wang, Rui Li, Chengjun Xie, Danfeng Hong, and Man Zhou. 2025. Pan-mamba: Effective pan-sharpening with state space model.Information Fusion115 (2025), 102779

    2025

  16. [24]

    Dan Hendrycks and Kevin Gimpel. 2016. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415(2016)

    2016 arXiv

  17. [25]

    Jie Hu, Li Shen, and Gang Sun. 2018. Squeeze-and-excitation networks. InProceed- ings of the IEEE conference on computer vision and pattern recognition. 7132–7141

    2018

  18. [26]

    Tao Huang, Xiaohuan Pei, Shan You, Fei Wang, Chen Qian, and Chang Xu. 2024. Localmamba: Visual state space model with windowed selective scan.arXiv preprint arXiv:2403.09338(2024)

    2024 arXiv

  19. [27]

    Rong Kang, Yue Cao, Mingsheng Long, Jianmin Wang, and Philip S Yu. 2019. Maximum-margin hamming hashing. InProceedings of the IEEE/CVF international conference on computer vision. 8252–8261

    2019

  20. [28]

    Alex Krizhevsky, Geoffrey Hinton, et al. 2009. Learning multiple layers of features from tiny images. (2009)

    2009

  21. [29]

    Brian Kulis and Trevor Darrell. 2009. Learning to hash with binary reconstructive embeddings. InProceedings of the Conference and Workshop on Neural Information Processing Systems. 1042–1050

    2009

  22. [30]

    Tao Li, Zheng Zhang, Lishen Pei, and Yan Gan. 2022. HashFormer: Vision transformer based deep hashing for image retrieval.IEEE Signal Processing Letters 29 (2022), 827–831

    2022

  23. [31]

    Wu-Jun Li, Sheng Wang, and Wang-Cheng Kang. 2016. Feature learning based deep supervised hashing with pairwise labels. InProceedings of the International Joint Conference on Artificial Intelligence. 1711–1717

    2016

  24. [32]

    Xue Li, Jiong Yu, Shaochen Jiang, Hongchun Lu, and Ziyang Li. 2023. Msvit: training multiscale vision transformers for image retrieval.IEEE Transactions on Multimedia(2023), 1 – 15

    2023

  25. [33]

    Guosheng Lin, Chunhua Shen, Qinfeng Shi, Anton Van den Hengel, and David Suter. 2014. Fast supervised hashing with decision trees for high-dimensional data. MambaHash: Visual State Space Deep Hashing Model for Large-Scale Image Retrieval Conference’17, July 2017, Washington, D...

    2014

  26. [34]

    Kevin Lin, Huei-Fang Yang, Jen-Hao Hsiao, and Chu-Song Chen. 2015. Deep learning of binary hash codes for fast image retrieval. InProceedings of the IEEE conference on computer vision and pattern recognition workshops. 27–35

    2015

  27. [35]

    Haomiao Liu, Ruiping Wang, Shiguang Shan, and Xilin Chen. 2016. Deep super- vised hashing for fast image retrieval. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2064–2072

    2016

  28. [36]

    Jiarun Liu, Hao Yang, Hong-Yu Zhou, Yan Xi, Lequan Yu, Cheng Li, Yong Liang, Guangming Shi, Yizhou Yu, Shaoting Zhang, et al. 2024. Swin-umamba: Mamba- based unet with imagenet-based pretraining. InInternational Conference on Med- ical Image Computing and Computer-Assisted Int...

    2024

  29. [37]

    Wei Liu, Jun Wang, Rongrong Ji, Yu-Gang Jiang, and Shih-Fu Chang. 2012. Su- pervised hashing with kernels. InProceedings of the IEEE conference on computer vision and pattern recognition. 2074–2081

    2012

  30. [38]

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, and Yunfan Liu. 2024. VMamba: Visual State Space Model.arXiv preprint arXiv:2401.10166(2024)

    2024 arXiv

  31. [39]

    Jun Ma, Feifei Li, and Bo Wang. 2024. U-mamba: Enhancing long-range de- pendency for biomedical image segmentation.arXiv preprint arXiv:2401.04722 (2024)

    2024 arXiv

  32. [40]

    Eric Nguyen, Karan Goel, Albert Gu, Gordon Downs, Preey Shah, Tri Dao, Stephen Baccus, and Christopher Ré. 2022. S4nd: Modeling images and videos as multidi- mensional signals with state spaces.Advances in neural information processing systems35 (2022), 2846–2861

    2022

  33. [41]

    Aude Oliva and Antonio Torralba. 2001. Modeling the shape of the scene: A holistic representation of the spatial envelope.International journal of computer vision42 (2001), 145–175

    2001

  34. [42]

    Namuk Park and Songkuk Kim. 2022. How do vision transformers work?. In Proceedings of the International Conference on Learning Representations

    2022

  35. [43]

    Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, et al

  36. [44]

    Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming- Hsuan Yang, and Fahad Shahbaz Khan. 2023. Swiftformer: Efficient additive attention for transformer-based real-time mobile vision applications. InProceed- ings of the IEEE/CVF International Conference on Co...

    2023

  37. [45]

    Zhengzhong Tu, Hossein Talebi, Han Zhang, Feng Yang, Peyman Milanfar, Alan Bovik, and Yinxiao Li. 2022. Maxvit: Multi-axis vision transformer. InEuropean conference on computer vision. Springer, 459–479

    2022

  38. [46]

    Chloe Wang, Oleksii Tsepa, Jun Ma, and Bo Wang. 2024. Graph-mamba: Towards long-range graph sequence modeling with selective state spaces.arXiv preprint arXiv:2402.00789(2024)

    2024 arXiv

  39. [47]

    Qilong Wang, Banggu Wu, Pengfei Zhu, Peihua Li, Wangmeng Zuo, and Qinghua Hu. 2020. ECA-Net: Efficient channel attention for deep convolutional neural networks. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition. 11534–11542

    2020

  40. [48]

    Shikui Wei, Lixin Liao, Jia Li, Qinjie Zheng, Fei Yang, and Yao Zhao. 2019. Saliency inside: Learning attentive CNNs for content-based image retrieval.IEEE Transac- tions on image processing28, 9 (2019), 4580–4593

    2019

  41. [49]

    Yair Weiss, Antonio Torralba, and Rob Fergus. 2008. Spectral hashing. InProceed- ings of the Conference and Workshop on Neural Information Processing Systems. 1753–1760

    2008

  42. [50]

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

    2018

  43. [51]

    Xingming Xiao, Shu Cao, Liejun Wang, Shuli Cheng, and Erdong Yuan. 2024. Deep hashing image retrieval based on hybrid neural network and optimized metric learning.Knowledge-Based Systems284 (2024), 111336

    2024

  44. [52]

    Saining Xie, Ross Girshick, Piotr Dollár, Zhuowen Tu, and Kaiming He. 2017. Aggregated residual transformations for deep neural networks. InProceedings of the IEEE conference on computer vision and pattern recognition. 1492–1500

    2017

  45. [53]

    Chenhongyi Yang, Zehui Chen, Miguel Espinosa, Linus Ericsson, Zhenyu Wang, Jiaming Liu, and Elliot J Crowley. 2024. Plainmamba: Improving non-hierarchical mamba in visual recognition.arXiv preprint arXiv:2403.17695(2024)

    2024 arXiv

  46. [54]

    Weihao Yu and Xinchao Wang. 2024. MambaOut: Do We Really Need Mamba for Vision?arXiv preprint arXiv:2405.07992(2024)

    2024 arXiv

  47. [55]

    Li Yuan, Tao Wang, Xiaopeng Zhang, Francis EH Tay, Zequn Jie, Wei Liu, and Jiashi Feng. 2020. Central similarity quantization for efficient image and video retrieval. InProceedings of the IEEE conference on computer vision and pattern recognition. 3083–3092

    2020

  48. [56]

    Dell Zhang, Jun Wang, Deng Cai, and Jinsong Lu. 2010. Self-taught hashing for fast similarity search. InProceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval. 18–25

    2010

  49. [57]

    Ting Zhang, Guo-Jun Qi, Bin Xiao, and Jingdong Wang. 2017. Interleaved group convolutions. InProceedings of the IEEE international conference on computer vision. 4373–4382

    2017

  50. [58]

    Zheng Zhang, Qin Zou, Yuewei Lin, Long Chen, and Song Wang. 2019. Improved deep hashing with soft pairwise similarity for multi-label image retrieval.IEEE Transactions on Multimedia22, 2 (2019), 540–553

    2019

  51. [59]

    Xiangtao Zheng, Yichao Zhang, and Xiaoqiang Lu. 2020. Deep balanced discrete hashing for image retrieval.Neurocomputing403 (2020), 224–236

    2020

  52. [60]

    Han Zhu, Mingsheng Long, Jianmin Wang, and Yue Cao. 2016. Deep hashing network for efficient similarity retrieval. InProceedings of the AAAI conference on artificial intelligence. 2415–2421

    2016

  53. [61]

    Lianghui Zhu, Bencheng Liao, Qian Zhang, Xinlong Wang, Wenyu Liu, and Xinggang Wang. 2024. Vision mamba: Efficient visual representation learning with bidirectional state space model.arXiv preprint arXiv:2401.09417(2024)

    2024 arXiv

  54. [62]

    Lei Zhu, Chaoqun Zheng, Weili Guan, Jingjing Li, Yang Yang, and Heng Tao Shen

  55. [2015]

    Imagenet large scale visual recognition challenge.International journal of computer vision115 (2015), 211–252

    2015

  56. [2019]

    In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Bag of tricks for image classification with convolutional neural networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 558–567

  57. [2023]

    Multi-modal hashing for efficient multimedia retrieval: A survey.IEEE Transactions on Knowledge and Data Engineering36, 1 (2023), 239–260

    2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.