Pith. sign in

REVIEW 3 major objections 6 minor 62 references

SpikeVideoFormer: An Efficient Spike-Driven Video Transformer with Hamming Attention and $\mathcal{O}(T)$ Complexity

T0 review · 3 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read SpikeVideoFormer claims Hamming attention between binary spikes outperforms dot-product attention in spike-driven video transformers, setting SNN records on three video tasks at linear temporal cost.

desk verdict Solid SNN-video transformer with a useful empirical result, but the abstract overclaims and the key ablation confounds Hamming score with threshold scaling. read the letter →

arxiv 2505.10352 v1 pith:NYJBV2XI submitted 2025-05-15 cs.CV cs.AIcs.LG

Shihao Zou , Qingfeng Li , Wei Ji , Jingjing Li , Yongkui Yang , Guoqi Li , Chao Dong This is my paper · ORCID
classification cs.CVcs.AIcs.LG
keywords spikingneuralnetworksvideotransformerHammingattentionlineartemporalcomplexityhumanposetrackingsemanticsegmentationbinaryembeddingenergy-efficientvision
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 establish that a spike-driven video transformer can be both accurate and efficient if the dot-product score in attention is replaced by normalized Hamming similarity between binary spike vectors. The authors argue that dot-product attention is a poor fit for 0/1 spikes, because a query containing zeros silently discards the corresponding key entries and can give identical scores to very different keys. They propose spike-driven Hamming attention (SDHA), whose score is a scaled dot product of bipolar vectors $(2Q_s-1)$ and $(2K_s-1)$, so it still runs in linear token complexity and can be implemented with additions and threshold shifts. On video classification, human pose tracking, and video semantic segmentation, the resulting model reports state-of-the-art accuracy among SNN methods, with over 15% gains on the latter two tasks, while matching recent ANN methods at a fraction of the power. If the claim holds, SNNs become a viable backbone for long video streams, not just single images.

What carries the argument

The central object is spike-driven Hamming attention (SDHA). Its score is the normalized Hamming similarity $f_H(q_s,k_s)=1-\frac{1}{D}\sum_i \mathbf{1}(q_s^{(i)}\ne k_s^{(i)})$, rewritten as $\frac{1}{2}+\frac{1}{2D}(2q_s-1)^\top(2k_s-1)$ so that 0/1 spikes become $\pm 1$ vectors by bit shifts, the $1/(2D)$ scale folds into the LIF neuron threshold, and the product is reordered to $(2Q_s-1)[(2K_s-1)^\top V_s]$, giving $\mathcal{O}(ND^2)$ complexity linear in token count. The justification comes from Proposition 3.1, which guarantees that with high probability $f_H$ tracks a monotone function of cosine similarity when the spikes are sign random projections of real features. Around SDHA, the paper constructs a space-time joint attention that reshapes the video tokens to length $TN$ before applying attention; this design beats hierarchical and factorized alternatives in its experiments while keeping temporal complexity $\mathcal{O}(T)$. The leaky integrate-and-fire neuron's membrane accumulation carries temporal encoding through the CNN stages.

What would settle it

Extract real spike queries $Q_s$ and keys $K_s$ from a trained SpikeVideoFormer together with the pre-threshold membrane features $Q,K$ at several layers, and measure the average error $|f_H(Q_s,K_s)-g(f_C(Q,K))|$ over many pairs at the channel widths actually used. If the error is large at those widths, Proposition 3.1 is not controlling what the network computes. A second check is ablative: keep the architecture identical but score attention with a centered dot product $(Q_s-\mu)^\top(K_s-\mu)$; if that recovers most of the gain, the contribution is zero-centering rather than Hamming geometry.

Watch

Extended reading notes

Core claim

At its core, the paper claims that normalized Hamming similarity, not dot-product, is the correct attention score for spike-driven transformers, and that this substitution is what lets an SNN transformer compete on video. The reported numbers are 79.8% top-1 on Kinetics-400 (against 75.5% for the prior spike-driven transformer), a PA-MPJPE of 39.8 mm on MMHPSD at $T=8$ (against 45.7 mm for the best prior transformer-based SNN), and 73.1% mIoU on CityScapes and 37.9% on VSPW (against 65.9% and 32.3% for prior SNN baselines). The theoretical bridge is Proposition 3.1, adapted from the binary-embedding Johnson–Lindenstrauss lemma: for binary embeddings $q_s=\mathrm{sign}(Aq)$ with Gaussian $A$, the normalized Hamming similarity $f_H(q_s,k_s)$ is within $\delta$ of $g(f_C(q,k))$, a monotone function of cosine similarity, with probability at least $1-2e^{-\delta^2 D}$. Since $g$ is monotone, the ranking of keys under Hamming similarity should match the ranking under cosine attention. The paper rewrites $f_H$ as $\frac{1}{2}+\frac{1}{2D}(2q_s-1)^\top(2k_s-1)$, which turns the computation into additions with a threshold-scaled spiking neuron and preserves the linear-complexity rearrangement of spike-driven attention. It also selects joint space-time attention—flattening the $T\times N$ tokens into one sequence before SDHA—as the best of three space-time designs, all of which run in $\mathcal{O}(TND^2)$.

Load-bearing premise

The load-bearing premise is that the 0/1 spike patterns emitted by the network's neurons behave as if they were sign random projections of real-valued feature vectors; the paper does not show that the actual leaky integrate-and-fire spikes satisfy this, and if they do not, the 'theoretically guided' label is an analogy and the case rests on the experiments.

Editorial extensions

If this is right

  • On Kinetics-400, the model reports 79.8% top-1 accuracy, 4.3 points above the strongest prior spike-driven transformer and within 0.8 points of Swin-B while using roughly one third of its power.
  • On MMHPSD human pose tracking, PA-MPJPE falls to 39.8 mm at $T=8$ and 47.5 mm at $T=32$; the $T=8$ result is within 0.1 mm of the ANN method GLoT while using far less power.
  • In video semantic segmentation, Hamming attention raises mIoU by 7.2 points on CityScapes and 5.6 points on VSPW over the best prior SNN transformer, and replacing Hamming with dot-product in the ablation costs 7.8 points of mIoU.
  • All three space-time attention designs (joint, hierarchical, factorized) run in $\mathcal{O}(TND^2)$, so the linear temporal complexity holds regardless of which design is chosen; the joint design is reported as both the simplest and the most accurate.
  • Power measurements based on 45nm energy constants place the model at roughly $\times16$, $\times10$, and $\times5$ efficiency gains over recent ANN methods on the three tasks, with event-stream inputs reducing power further because of lower spiking rates.

Reading between the lines

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

  • Because the theoretical guarantee in Proposition 3.1 applies to sign random Gaussian projections of real vectors, and LIF spike trains are not shown to be such projections, the ranking-preservation argument is, at best, an analogy for the actual network; the empirical ablation carries the case.
  • The rewritten Hamming score is mathematically a centered bipolar dot product, so a plausible alternative explanation of the gains is that subtracting a reference level removes the all-zeros collapse of raw spike dot products; this suggests a cheap test: keep dot-product attention but center the spikes before scoring.
  • Because the temporal cost is $\mathcal{O}(T)$ and event streams are naturally sparse, deploying this architecture on neuromorphic hardware for online, long-horizon video tasks is a direct next step; end-to-end latency and energy on such hardware would test whether the simulated power advantages survive.
  • The finding that joint space-time attention beats decomposed designs mirrors what the paper cites from video generation, so SDHA may transfer to spike-driven video diffusion or autoregressive generation, where space-time joint attention is already favored.
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

3 major / 6 minor

Summary. The paper proposes SpikeVideoFormer, a spike-driven video transformer with linear temporal complexity O(T). The core contribution is spike-driven Hamming attention (SDHA), which replaces the dot-product score in prior spike-driven self-attention with a normalized Hamming similarity, motivated by a Johnson–Lindenstrauss-type result on binary embeddings. The authors further compare three space-time attention designs (joint, hierarchical, factorized) and select joint attention, which they show empirically performs best. The model is evaluated on Kinetics-400 video classification, MMHPSD human pose tracking (with video and event-stream inputs), and CityScapes/VSPW video semantic segmentation. The paper reports state-of-the-art performance among SNN baselines on these tasks, with large efficiency gains over ANN video models, and it includes a GitHub repository with code.

Significance. If the central claims hold, this is a useful and timely contribution: it is, to my knowledge, among the first spike-driven video transformers, it demonstrates a credible efficiency-performance trade-off across three diverse video tasks, and the space-time attention design comparison provides practical guidance. The paper ships reproducible code, detailed training hyperparameters in the appendices, and energy estimates consistent with the SNN literature. The theoretical motivation via binary embedding is attractive, although, as discussed below, its applicability to trained LIF features is not established. The empirical gains over strong SNN baselines are substantial in several configurations, especially on video semantic segmentation.

major comments (3)
  1. [Abstract and §4.2, Tables 3 and 5] The abstract's claim of 'over 15% improvement on the latter two tasks' is not supported by the tables. For human pose tracking on MMHPSD (Table 3), the best improvement in PA-MPJPE over the strongest SNN baseline (Meta-SpikeFormer*, T=32, video input) is (45.7 - 39.8)/45.7 = 12.9%; the T=8 improvement is about 9.6%. For video semantic segmentation, the improvement over Meta-SpikeFormer is 10.9% on CityScapes (65.9 to 73.1 mIoU) and 17.3% on VSPW (32.3 to 37.9 mIoU). Thus the 'over 15%' claim is not representative of the pose task and is dataset-dependent for segmentation. The abstract and the ablation discussion in §4.4 (which states '15% in pose tracking') should be corrected to precise, table-consistent numbers.
  2. [§4.4, Table 6] The ablation that isolates the attention score function is confounded with the threshold scale. The row 'Att. Score Ham→Dot-prod' reports PA-MPJPE 45.7 and mIoU 65.9, which are exactly the Meta-SpikeFormer* results in Tables 3 and 4, indicating that this row uses the dot-product SDSA with its original threshold scale s=1/8, whereas the proposed SDHA uses s=1/(2D). The threshold-scale ablation in the same table shows that Hamming attention at s=1/8 gives PA-MPJPE 46.7, which is worse than the dot-product row's 45.7, and only mIoU improves (68.5 vs 65.9). Therefore the current evidence does not establish that Hamming similarity itself outperforms dot-product at a matched scale; the observed gains could be due primarily to the better threshold scale. Since Proposition 3.1 does not apply directly to learned LIF features (see next comment), this ablation is the main empirical support for the central novelty. The authors should add a controlled comparison with both score functions evaluated at several threshold scales on both tasks, and report these numbers in the main text.
  3. [§3.3, Proposition 3.1] The theoretical motivation is presented as a proposition about real-valued vectors q,k and their binary embeddings qs=sign(Aq), ks=sign(Ak) with A having i.i.d. Gaussian entries. However, the queries and keys used in SpikeVideoFormer are produced by learned convolutions, batch normalization, and LIF spiking; the paper provides no argument or measurement that these spikes behave like sign(random Gaussian projections) of the corresponding real features. As stated, the proposition is a known result about binary embeddings and does not directly guarantee anything about the learned spike features. The paper should (a) explicitly state that Proposition 3.1 is an analogy/motivation rather than a theorem about the trained model, or (b) provide a quantitative verification on actual trained features, e.g., by comparing f_H with g(f_C) for the model's Qs and Ks. This is load-bearing because without such a qualification the phrase 'theoretically guided adaptation' overstates what is proven, and the empirical ablation then becomes the sole support for the proposed attention mechanism.
minor comments (6)
  1. [§3.3, Proposition 3.1 and Appendix A] There is a dimension/notation inconsistency: the proposition states A∈R^{C×D} while the product Aq requires A∈R^{D×C}; Appendix A correctly uses A∈R^{D×C}. Also, 'sign' typically outputs ±1, but qs,ks are defined as {0,1}^D; the mapping from sign outputs to binary values should be made explicit.
  2. [§3.3, Eq. (10)] The notation SN_{2D} is ambiguous: it is unclear whether the subscript denotes a threshold scale 2D·u_th, a neuron index, or a number of steps. Please define the subscript convention when the threshold is scaled.
  3. [§4.4, Table 6] The table headers for the two 'Power (mJ)' columns do not indicate which task they refer to. Since each row reports pose and VSS numbers interleaved, the reader cannot tell which power value belongs to which task without inspecting the text.
  4. [Appendix G] There is a typo: 'calssification head' should be 'classification head'.
  5. [§2, Visual Cognitive Neuroscience paragraph] The paragraph on visual cognitive neuroscience is disconnected from the rest of the related work and does not clearly motivate the method. Consider removing it or integrating it with the SNN motivation in the introduction.
  6. [§4.4, first paragraph] The sentence describing the ablation results says '15% in pose tracking and 10% in VSS'; these percentages should be updated to match the corrected relative improvements (approximately 13% and 11% respectively for the numbers in Table 6).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SDHA is an independent construction; the JL-based motivation is an external theorem and the threshold scale is a transparent design choice.

full rationale

The central claim—replacing dot-product attention with normalized Hamming similarity in a spike-driven transformer—is not equivalent to its inputs by construction. Proposition 3.1 is a known JL-type binary-embedding result (Jacques et al., 2013; Yi et al., 2015) whose proof is reproduced in Appendix A, and it is cited as a motivation rather than as a definition of the target result. The actual LIF spike features are not shown to satisfy the sign(random-Gaussian-projection) premise, so the theoretical link is an analogy; that is a generalization/validity concern, not a circular reduction. The 1/(2D) scaling in Eq. (10) is algebraically inherited from the normalized-Hamming definition (Eq. (8)) and is tested against alternatives in Table 6; it is a hyperparameter choice, not a fitted prediction. The Ham→Dot-prod ablation does vary both the score function and the threshold scale, which is a legitimate experimental confound, but it does not make the derived quantity equal to the input by definition. Self-citations (e.g., Meta-SpikeFormer, Yao et al., 2024a) supply baselines and architectural settings but are not the load-bearing justification for SDHA. No circular step can be exhibited, so the circularity score is 0. Separately, the abstract's 'over 15% improvement' claim is not fully supported by Tables 3 and 4, but that is a reporting-accuracy issue, not a circularity.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central method relies on established LIF modeling and a known JL-type theorem, with the main free choices being the threshold scale and task-specific loss weights. No new physical entities are introduced.

free parameters (2)
  • threshold scale s in SDHA = 1/2D per layer
    The scaling factor is set to 1/2D based on the algebraic form of fH, but the ablation in Table 6 shows it outperforms fixed alternatives (1/8, 1/64, 1/512, 1/2048), indicating it is partially empirically selected.
  • loss weights lambda_pose, lambda_shape, lambda_trans, lambda_3D, lambda_2D = not stated in the paper
    The total pose-tracking loss in Appendix H includes five weights that are never specified; their values affect the reported pose metrics and are not reproducible from the text.
assumptions (3)
  • standard math JL Lemma on binary embeddings (Proposition 3.1)
    The theoretical motivation for Hamming attention is the Johnson-Lindenstrauss-type guarantee that binary embeddings preserve cosine similarity up to a monotone transform. This is a known theorem (Jacques et al., 2013; Yi et al., 2015), but its application assumes the learned spike queries/keys behave like sign-random-projections of real features.
  • domain assumption LIF neuron model with threshold-based spiking
    The architecture assumes the standard Leaky Integrate-and-Fire model (Sec. 3.1) with surrogate gradient training, a common assumption in the SNN literature.
  • domain assumption Energy model: SNN power = rho * FLOPs * E_MAC with E_MAC=4.6pJ and E_AC=0.9pJ
    Power consumption claims in Sec. 4 and Appendix E use theoretical FLOP counts and fixed per-operation energy from Horowitz (2014); they are not validated on actual neuromorphic hardware.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SpikeVideoFormer: An Efficient Spike-Driven Video Transformer with Hamming Attention and $\mathcal{O}(T)$ Complexity." pith.science (2026). https://pith.science/paper/NYJBV2XI

@misc{pith2026250510352,
  author       = {Pith},
  title        = {Pith review of: SpikeVideoFormer: An Efficient Spike-Driven Video Transformer with Hamming Attention and $\mathcalO(T)$ Complexity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NYJBV2XI}},
  note         = {Machine review of arXiv:2505.10352}
}
abstract

Spiking Neural Networks (SNNs) have shown competitive performance to Artificial Neural Networks (ANNs) in various vision tasks, while offering superior energy efficiency. However, existing SNN-based Transformers primarily focus on single-image tasks, emphasizing spatial features while not effectively leveraging SNNs' efficiency in video-based vision tasks. In this paper, we introduce SpikeVideoFormer, an efficient spike-driven video Transformer, featuring linear temporal complexity $\mathcal{O}(T)$. Specifically, we design a spike-driven Hamming attention (SDHA) which provides a theoretically guided adaptation from traditional real-valued attention to spike-driven attention. Building on SDHA, we further analyze various spike-driven space-time attention designs and identify an optimal scheme that delivers appealing performance for video tasks, while maintaining only linear temporal complexity. The generalization ability and efficiency of our model are demonstrated across diverse downstream video tasks, including classification, human pose tracking, and semantic segmentation. Empirical results show our method achieves state-of-the-art (SOTA) performance compared to existing SNN approaches, with over 15\% improvement on the latter two tasks. Additionally, it matches the performance of recent ANN-based methods while offering significant efficiency gains, achieving $\times 16$, $\times 10$ and $\times 5$ improvements on the three tasks. https://github.com/JimmyZou/SpikeVideoFormer

Figures

Figures reproduced from arXiv: 2505.10352 by the authors.

Figure 1
Figure 1. Architecture of our proposed SpikeVideoFormer. An input video with a shape of T × H × W × 3 undergoes temporal spiking over time, after which it passes through two spike-driven CNN blocks, along with downsample modules. This is followed by two spike-driven spatiotemporal transformers, also accompanied by downsample modules. Finally, the extracted video features are forwarded to the regression or classification head … view at source ↗
Figure 2
Figure 2. Intuitive comparison of attention scores between spike query and keys using Dot-product (Dot.) and normalized Hamming similarity (Ham.). When the spike query contains no elements, the dot-product ignores the corresponding elements in the spike keys, resulting in identical scores for four distinct spike keys as examples. This illustrates the dot-product’s limitation in accurately capturing the similarity between bina… view at source ↗
Figure 4
Figure 4. Space-time Spike-Driven Attention Designs [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (6 more)
Figure 5
Figure 5. Figure 5: Architecture Details of three different spike-driven space-time attentions. normalization and spiking layers to convert to spike query QS, key KS and value VS. They will be re￾shaped to B × T N × D and forwarded to SDHA for spatiotemporal feature fusion. The output fin…
Figure 6
Figure 6. Figure 6: Workflow of human pose tracking using our proposed SpikeVideoFormer. The regression head consists of three parallel linear layers to regress the SMPL shape parameters. rameters, global translations, and 3D and 2D joint positions to train our model. The total loss funct…
Figure 7
Figure 7. Figure 7: Qualitative Results of Our SpikeVideoFormer on Human Pose Tracking. Additional Implementation Details. The input dimensions for both modalities are T × 256 × 256 × 3. The event stream is evenly split into 3T segments over time, with each segment transformed into an eve…
Figure 9
Figure 9. Figure 9: Pipeline of video semantic segmentation using SpikeV￾ideoFormer as encoder, spike-driven Hamming attention as Mem￾ory Read and Fusion module, and SpikeFPN as decoder [PITH_FULL_IMAGE:figures/full_fig_p017_9.png]
Figure 8
Figure 8. Figure 8: Visualization of Attention Maps in Human Pose Tracking of features serve as the key and value for cross-attention via SDHA. The fused features of the last step across all four scales are subsequently used as pyramid features in the decoder SpikeFPN, which predicts pixe…
Figure 10
Figure 10. Figure 10: Qualitative Results of Our SpikeVideoFormer on VSS. 17 [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 50 canonical work pages

  1. [1]

    Vivit: A video vision transformer

    Arnab, A., Dehghani, M., Heigold, G., Sun, C., Lu c i \'c , M., and Schmid, C. Vivit: A video vision transformer. In ICCV, 2021

    2021

  2. [2]

    Object domain and modality in the ventral visual pathway

    Bi, Y., Wang, X., and Caramazza, A. Object domain and modality in the ventral visual pathway. Trends in cognitive sciences, 2016

    work page 2016

  3. [3]

    Stable video diffusion: Scaling latent video diffusion models to large datasets

    Blattmann, A., Dockhorn, T., Kulal, S., Mendelevitch, D., Kilian, M., Lorenz, D., Levi, Y., English, Z., Voleti, V., Letts, A., et al. Stable video diffusion: Scaling latent video diffusion models to large datasets. arXiv preprint arXiv:2311.15127, 2023

    arXiv 2023

  4. [4]

    Spiking diffusion models

    Cao, J., Guo, H., Wang, Z., Zhou, D., Cheng, H., Zhang, Q., and Xu, R. Spiking diffusion models. IEEE TAI, 2024

    work page 2024

  5. [5]

    Chen, L.-C., Papandreou, G., Kokkinos, I., Murphy, K., and Yuille, A. L. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE TPAMI, 2017

    work page 2017

  6. [6]

    Xception: Deep learning with depthwise separable convolutions

    Chollet, F. Xception: Deep learning with depthwise separable convolutions. In CVPR, 2017

    work page 2017

  7. [7]

    The cityscapes dataset for semantic urban scene understanding

    Cordts, M., Omran, M., Ramos, S., Rehfeld, T., Enzweiler, M., Benenson, R., Franke, U., Roth, S., and Schiele, B. The cityscapes dataset for semantic urban scene understanding. In CVPR, 2016

    work page 2016

  8. [8]

    Dapello, J., Marques, T., Schrimpf, M., Geiger, F., Cox, D., and DiCarlo, J. J. Simulating a primary visual cortex at the front of cnns improves robustness to image perturbations. NeurIPS, 2020

    work page 2020

Show all 62 references
  1. [9]

    Imagenet: A large-scale hierarchical image database

    Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., and Fei-Fei, L. Imagenet: A large-scale hierarchical image database. In CVPR, 2009

    2009

  2. [10]

    Repvgg: Making vgg-style convnets great again

    Ding, X., Zhang, X., Ma, N., Han, J., Ding, G., and Sun, J. Repvgg: Making vgg-style convnets great again. In CVPR, 2021

    2021

  3. [11]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. An image is worth 16x16 words: Transformers for image recognition at scale. ICLR, 2021

    2021

  4. [12]

    Deep residual learning in spiking neural networks

    Fang, W., Yu, Z., Chen, Y., Huang, T., Masquelier, T., and Tian, Y. Deep residual learning in spiking neural networks. NeurIPS, 2021

    2021

  5. [13]

    Bottom-up and top-down approaches for the design of neuromorphic processing systems: tradeoffs and synergies between natural and artificial intelligence

    Frenkel, C., Bol, D., and Indiveri, G. Bottom-up and top-down approaches for the design of neuromorphic processing systems: tradeoffs and synergies between natural and artificial intelligence. Proceedings of the IEEE, 2023

    2023

  6. [14]

    Deep residual learning for image recognition

    He, K., Zhang, X., Ren, S., and Sun, J. Deep residual learning for image recognition. In CVPR, 2016

    2016

  7. [15]

    1.1 computing's energy problem (and what we can do about it)

    Horowitz, M. 1.1 computing's energy problem (and what we can do about it). In ISSCC, 2014

    2014

  8. [16]

    Fast-snn: Fast spiking neural network by converting quantized ann

    Hu, Y., Zheng, Q., Jiang, X., and Pan, G. Fast-snn: Fast spiking neural network by converting quantized ann. IEEE TPAMI, 2023

    2023

  9. [17]

    Advancing spiking neural networks toward deep residual learning

    Hu, Y., Deng, L., Wu, Y., Yao, M., and Li, G. Advancing spiking neural networks toward deep residual learning. IEEE TNNLS, 2024

    2024

  10. [18]

    N., Boufounos, P

    Jacques, L., Laska, J. N., Boufounos, P. T., and Baraniuk, R. G. Robust 1-bit compressive sensing via binary stable embeddings of sparse vectors. IEEE Transactions on Information Theory, 2013

    2013

  11. [19]

    Fully spiking variational autoencoder

    Kamata, H., Mukuta, Y., and Harada, T. Fully spiking variational autoencoder. In AAAI, 2022

    2022

  12. [20]

    The kinetics human action video dataset

    Kay, W., Carreira, J., Simonyan, K., Zhang, B., Hillier, C., Vijayanarasimhan, S., Viola, F., Green, T., Back, T., Natsev, P., et al. The kinetics human action video dataset. arXiv preprint arXiv:1705.06950, 2017

    2017 arXiv

  13. [21]

    Kocabas, M., Athanasiou, N., and Black, M. J. Vibe: Video inference for human body pose and shape estimation. In CVPR, 2020

    2020

  14. [22]

    M., Pandit, T., Merkel, C., Kubendran, R., Aimone, J

    Kudithipudi, D., Schuman, C., Vineyard, C. M., Pandit, T., Merkel, C., Kubendran, R., Aimone, J. B., Orchard, G., Mayr, C., and Benosman, R. Neuromorphic computing at scale. Nature, 2025

    2025

  15. [23]

    Brain-inspired computing: A systematic survey and future trends

    Li, G., Deng, L., Tang, H., Pan, G., Tian, Y., Roy, K., and Maass, W. Brain-inspired computing: A systematic survey and future trends. Proceedings of the IEEE, 2024

    2024

  16. [24]

    Firefly: A high-throughput hardware accelerator for spiking neural networks with efficient dsp and memory optimization

    Li, J., Shen, G., Zhao, D., Zhang, Q., and Zeng, Y. Firefly: A high-throughput hardware accelerator for spiking neural networks with efficient dsp and memory optimization. IEEE TVLSI, 2023 a

    2023

  17. [25]

    Spikeclip: A contrastive language-image pretrained spiking neural network

    Li, T., Liu, W., Lv, C., Gu, Y., Xu, J., Zhang, C., Wu, M., Zheng, X., and Huang, X. Spikeclip: A contrastive language-image pretrained spiking neural network. arXiv preprint arXiv:2310.06488, 2023 b

    2023 arXiv

  18. [26]

    Video swin transformer

    Liu, Z., Ning, J., Cao, Y., Wei, Y., Zhang, Z., Lin, S., and Hu, H. Video swin transformer. In CVPR, 2022

    2022

  19. [27]

    Fully convolutional networks for semantic segmentation

    Long, J., Shelhamer, E., and Darrell, T. Fully convolutional networks for semantic segmentation. IEEE TPAMI, 2017

    2017

  20. [28]

    Loper, M., Mahmood, N., Romero, J., Pons-Moll, G., and Black, M. J. Smpl: A skinned multi-person linear model. ACM TOG, 2015

    2015

  21. [29]

    Integer-valued training and spike-driven inference spiking neural network for high-performance and energy-efficient object detection

    Luo, X., Yao, M., Chou, Y., Xu, B., and Li, G. Integer-valued training and spike-driven inference spiking neural network for high-performance and energy-efficient object detection. ECCV, 2024

    2024

  22. [30]

    Networks of spiking neurons: the third generation of neural network models

    Maass, W. Networks of spiking neurons: the third generation of neural network models. Neural networks, 1997

    1997

  23. [31]

    Vspw: A large-scale dataset for video scene parsing in the wild

    Miao, J., Wei, Y., Wu, Y., Liang, C., Li, G., and Yang, Y. Vspw: A large-scale dataset for video scene parsing in the wild. In CVPR, 2021

    2021

  24. [32]

    and Torralba, A

    Oliva, A. and Torralba, A. The role of context in object recognition. Trends in cognitive sciences, 2007

    2007

  25. [33]

    Local memory attention for fast video semantic segmentation

    Paul, M., Danelljan, M., Van Gool, L., and Timofte, R. Local memory attention for fast video semantic segmentation. In IROS, 2021

    2021

  26. [34]

    The human imagination: the cognitive neuroscience of visual mental imagery

    Pearson, J. The human imagination: the cognitive neuroscience of visual mental imagery. Nature reviews neuroscience, 2019

    2019

  27. [35]

    Mobilenetv2: Inverted residuals and linear bottlenecks

    Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.-C. Mobilenetv2: Inverted residuals and linear bottlenecks. In CVPR, 2018

    2018

  28. [36]

    Global-to-local modeling for video-based 3d human pose and shape estimation

    Shen, X., Yang, Z., Wang, X., Ma, J., Zhou, C., and Yang, Y. Global-to-local modeling for video-based 3d human pose and shape estimation. In CVPR, 2023

    2023

  29. [37]

    F., Klink, P

    Storm, J. F., Klink, P. C., Aru, J., Senn, W., Goebel, R., Pigorini, A., Avanzini, P., Vanduffel, W., Roelfsema, P. R., Massimini, M., et al. An integrative, multiscale view on neural theories of consciousness. Neuron, 2024

    2024

  30. [38]

    Deep directly-trained spiking neural networks for object detection

    Su, Q., Chou, Y., Hu, Y., Li, J., Mei, S., Zhang, Z., and Li, G. Deep directly-trained spiking neural networks for object detection. In ICCV, 2023

    2023

  31. [39]

    Multi-scale full spike pattern for semantic segmentation

    Su, Q., He, W., Wei, X., Xu, B., and Li, G. Multi-scale full spike pattern for semantic segmentation. Neural Networks, 2024

    2024

  32. [40]

    and Egner, T

    Summerfield, C. and Egner, T. Expectation (and attention) in visual cognition. Trends in cognitive sciences, 2009

    2009

  33. [41]

    Learning local and global temporal contexts for video semantic segmentation

    Sun, G., Liu, Y., Ding, H., Wu, M., and Van Gool, L. Learning local and global temporal contexts for video semantic segmentation. IEEE TPAMI, 2024

    2024

  34. [42]

    N., Kaiser, ., and Polosukhin, I

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, ., and Polosukhin, I. Attention is all you need. In NeurIPS, 2017

    2017

  35. [43]

    How brains beware: neural mechanisms of emotional attention

    Vuilleumier, P. How brains beware: neural mechanisms of emotional attention. Trends in cognitive sciences, 2005

    2005

  36. [44]

    Pssd-transformer: Powerful sparse spike-driven transformer for image semantic segmentation

    Wang, H., Liang, X., Zhang, T., Gu, Y., and Geng, W. Pssd-transformer: Powerful sparse spike-driven transformer for image semantic segmentation. In ACM MM, 2024

    2024

  37. [45]

    and Torresani, L

    Wang, J. and Torresani, L. Deformable video transformer. In CVPR, 2022

    2022

  38. [46]

    Non-local neural networks

    Wang, X., Girshick, R., Gupta, A., and He, K. Non-local neural networks. In CVPR, 2018

    2018

  39. [47]

    End-to-end video instance segmentation with transformers

    Wang, Y., Xu, Z., Wang, X., Shen, C., Cheng, B., Shen, H., and Xia, H. End-to-end video instance segmentation with transformers. In CVPR, 2021

    2021

  40. [48]

    Resnet strikes back: An improved training procedure in timm

    Wightman, R., Touvron, H., and J \'e gou, H. Resnet strikes back: An improved training procedure in timm. arXiv preprint arXiv:2110.00476, 2021

    2021 arXiv

  41. [49]

    M., and Luo, P

    Xie, E., Wang, W., Yu, Z., Anandkumar, A., Alvarez, J. M., and Luo, P. Segformer: Simple and efficient design for semantic segmentation with transformers. In NeurIPS, 2021

    2021

  42. [50]

    Reevaluating the sensory account of visual working memory storage

    Xu, Y. Reevaluating the sensory account of visual working memory storage. Trends in Cognitive Sciences, 2017

    2017

  43. [51]

    Cogvideox: Text-to-video diffusion models with an expert transformer

    Yang, Z., Teng, J., Zheng, W., Ding, M., Huang, S., Xu, J., Yang, Y., Hong, W., Zhang, X., Feng, G., et al. Cogvideox: Text-to-video diffusion models with an expert transformer. arXiv preprint arXiv:2408.06072, 2024

    2024 arXiv

  44. [52]

    Attention spiking neural networks

    Yao, M., Zhao, G., Zhang, H., Hu, Y., Deng, L., Tian, Y., Xu, B., and Li, G. Attention spiking neural networks. IEEE TPAMI, 2023

    2023

  45. [53]

    Spike-driven transformer v2: Meta spiking neural network architecture inspiring the design of next-generation neuromorphic chips

    Yao, M., Hu, J., Hu, T., Xu, Y., Zhou, Z., Tian, Y., Xu, B., and Li, G. Spike-driven transformer v2: Meta spiking neural network architecture inspiring the design of next-generation neuromorphic chips. ICLR, 2024 a

    2024

  46. [54]

    Spike-driven transformer

    Yao, M., Hu, J., Zhou, Z., Yuan, L., Tian, Y., Xu, B., and Li, G. Spike-driven transformer. NeurIPS, 2024 b

    2024

  47. [55]

    Yao, M., Richter, O., Zhao, G., Qiao, N., Xing, Y., Wang, D., Hu, T., Fang, W., Demirci, T., and Marchi, M. D. Spike-based dynamic computing with asynchronous sensing-computing neuromorphic chip. Nature Communications, 2025

    2025

  48. [56]

    Binary embedding: Fundamental limits and fast algorithm

    Yi, X., Caramanis, C., and Price, E. Binary embedding: Fundamental limits and fast algorithm. In ICML, 2015

    2015

  49. [57]

    Metaformer is actually what you need for vision

    Yu, W., Luo, M., Zhou, P., Si, C., Zhou, Y., Wang, X., Feng, J., and Yan, S. Metaformer is actually what you need for vision. In CVPR, 2022

    2022

  50. [58]

    Pyramid scene parsing network

    Zhao, H., Shi, J., Qi, X., Wang, X., and Jia, J. Pyramid scene parsing network. In CVPR, 2017

    2017

  51. [59]

    Direct training high-performance deep spiking neural networks: a review of theories and methods

    Zhou, C., Zhang, H., Yu, L., Ye, Y., Zhou, Z., Huang, L., Ma, Z., Fan, X., Zhou, H., and Tian, Y. Direct training high-performance deep spiking neural networks: a review of theories and methods. Frontiers in Neuroscience, 2024

    2024

  52. [60]

    Spikformer: When spiking neural network meets transformer

    Zhou, Z., Zhu, Y., He, C., Wang, Y., Yan, S., Tian, Y., and Yuan, L. Spikformer: When spiking neural network meets transformer. In ICLR, 2022

    2022

  53. [61]

    Eventhpe: Event-based 3d human pose and shape estimation

    Zou, S., Guo, C., Zuo, X., Wang, S., Wang, P., Hu, X., Chen, S., Gong, M., and Cheng, L. Eventhpe: Event-based 3d human pose and shape estimation. In ICCV, 2021

    2021

  54. [62]

    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 gl...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.