Pith. sign in

REVIEW 5 major objections 6 minor 44 references

Neuromorphic Sequential Arena: A Benchmark for Neuromorphic Temporal Processing

T0 review · 5 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper introduces the Neuromorphic Sequential Arena, a seven-task benchmark for spiking networks, and claims its Segregated Temporal Probe shows existing neuromorphic benchmarks can be solved without temporal memory while its tasks…

desk verdict Useful SNN benchmark with a load-bearing validity claim that needs rework before the 'more effective than existing benchmarks' conclusion can be trusted. read the letter →

arxiv 2505.22035 v1 pith:BXU442WT submitted 2025-05-28 cs.NE

Xinyi Chen , Chenxiang Ma , Yujie Wu , Kay Chen Tan , Jibin Wu This is my paper · ORCID
classification cs.NE
keywords spikingneuralnetworksneuromorphictemporalprocessingcreditassignmentbackpropagationthroughtimesequencebenchmarkingevent-basedvisionenergyefficiencyarchitecturecomparison
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper introduces the Neuromorphic Sequential Arena (NSA), a benchmark of seven real-world temporal processing tasks for spiking neural networks, spanning robot localization, activity recognition, EEG motor imagery, sound localization, lip-reading, audio denoising, and Mandarin speech recognition. Its central claim is that NSA measures temporal processing capacity more effectively than existing neuromorphic benchmarks, which are either static images repeated over time, event-based vision with limited motion, or keyword spotting solvable with short-term features. The evidence is the Segregated Temporal Probe (STP): training three variants of the same network—full spatio-temporal backpropagation, backpropagation without temporal gradients, and processing each time step independently—and comparing performance. On all seven NSA tasks, removing temporal propagation produces large performance drops, while on twelve common neuromorphic benchmarks the same ablation changes little on static and event-based vision and only moderately on keyword spotting. The paper also reports baselines for neuron models and architectures on accuracy, training speed, memory, and energy, concluding that high temporal performance and low cost rarely come together.

What carries the argument

The central object is the Segregated Temporal Probe (STP), a diagnostic that trains the same spiking network under three regimes: STBP, which keeps temporal recurrence in both forward and backward passes; SDBP, which keeps the forward recurrence but blocks error propagation across time steps; and NoTD, which removes the recurrence entirely so each time step is processed independently. Comparing the three accuracies isolates how much of a task's difficulty comes from temporal dependencies rather than spatial pattern matching. The probe is built on the standard leaky integrate-and-fire neuron, whose membrane potential update $u_l[t] = \lambda u_l[t-1](1-s_l[t-1]) + W^l s^{l-1}[t]$ is the forward temporal pathway that SDBP and NoTD sever in different ways. STP's role is to validate each benchmark task before trusting it: a task is temporally meaningful only if both ablated regimes clearly underperform full STBP.

What would settle it

Train the same NSA tasks under SDBP with gradient normalization or a matched learning-rate schedule so that the truncated backward pass has comparable gradient magnitudes to STBP; if the SDBP-to-STBP gap disappears under compensation, the STP conclusion that these tasks require full backward temporal credit assignment would be undercut. A complementary check is to shuffle the input frames on the AL task: if NoTD accuracy stays at chance on shuffled inputs while STBP accuracy drops, temporal order is what matters.

Watch

Extended reading notes

Core claim

NSA is intended as an effective, versatile, application-oriented benchmark for neuromorphic temporal processing. The authors' core discovery, in their own framing, is that a task's usefulness for temporal-processing research can be audited by the Segregated Temporal Probe, and by that audit the seven NSA tasks genuinely require temporal credit assignment while twelve commonly used neuromorphic benchmarks largely do not. Under full spatio-temporal backpropagation, a leaky integrate-and-fire network on NSA reaches 63.52% accuracy on autonomous localization and 81.27% on human activity recognition; cutting backward temporal propagation drops these by about five points, and removing temporal propagation entirely drops them by about ten and thirteen points respectively, with qualitatively similar degradation on the other five tasks. The same probe shows static and event-based vision benchmarks barely change when temporal propagation is removed, and keyword-spotting benchmarks show only marginal gaps between full and spatially-only backpropagation, which the authors attribute to short-term feature integration rather than genuine temporal dependency.

Load-bearing premise

The STP diagnostic assumes that the performance gap between STBP, SDBP, and NoTD is caused solely by removing temporal credit assignment, not by optimization side effects such as changes in gradient variance or optimization landscape.

Editorial extensions

If this is right

  • Research groups can use NSA as a shared testbed to compare spiking neuron models and architectures on the same seven tasks, with public baselines for accuracy, training speed, memory, and energy.
  • Claims that a new spiking network design improves temporal processing should be backed by tasks that pass the STP audit; otherwise the improvement may reflect spatial or short-term pattern matching.
  • Efficiency results indicate that parallel training schemes such as sliding parallel neurons, spiking state-space models, and gated spiking units speed up training roughly threefold over serial models, but often at higher memory and energy cost, so future designs need to optimize the trade-off.
  • The per-task rankings show no single neuron model or architecture wins everywhere: PMSN leads overall in average rank, GSN does best on denoising and speech recognition, and state-space spiking models handle noisy tasks well, suggesting task diversity matters for benchmarking.

Reading between the lines

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

  • STP could be applied to any new dataset before it is adopted as an SNN benchmark; a cheap three-run ablation would reveal whether temporal credit assignment is actually required, independent of the rest of NSA.
  • The same audit idea transfers to non-spiking sequence models: cutting temporal gradients in an RNN or temporal convolution would expose whether a benchmark rewards temporal structure or just per-timestep features.
  • The very low STBP accuracy on the sound localization task (8.88%) suggests NSA's tasks differ widely in difficulty, so future evaluations should report both absolute scores and the STP gap, not one without the other.
  • NSA's energy accounting uses a fixed per-operation cost from a 45 nm CMOS process; extending the same protocol to analog or in-memory neuromorphic hardware would make the efficiency comparisons more directly actionable.
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

5 major / 6 minor

Summary. This paper introduces the Neuromorphic Sequential Arena (NSA), a benchmark suite of seven tasks (AL, HAR, EEG-MI, SSL, ALR, AD, ASR) intended to evaluate the temporal processing capacity of spiking neural networks. To justify the benchmark's design, the authors propose the Segregated Temporal Probe (STP), which compares three training algorithms—STBP, SDBP, and NoTD—and interprets large performance gaps between them as evidence that a task contains rich temporal dependencies. Using STP, they argue that twelve existing neuromorphic benchmarks are inadequate, whereas NSA tasks are effective. They then provide extensive baselines for five spiking neuron models and five neural architectures, including accuracy, training speed, memory usage, and energy efficiency, and identify PMSN and Binary S4D/GSU as strong performers.

Significance. The paper addresses an important need in the neuromorphic SNN community for standardized, multi-task temporal processing benchmarks, and it delivers a substantial empirical comparison with openly released code and detailed hyperparameter tables. The efficiency analysis framework, though preliminary, is a useful addition. However, the central validation claim rests on the STP diagnostic, which is confounded: the probe does not isolate temporal credit assignment from optimization and capacity changes, and the paper's own supplementary results show that the training-algorithm gaps can appear in directions opposite to the claimed rule. The benchmark itself may be a valuable community resource, but the evidence that it is 'more effective' than existing benchmarks is not yet conclusive. The resource and baseline results are significant; the validation methodology needs to be strengthened.

major comments (5)
  1. [Sec. 2.3, Eqs. (1)-(6), Table 2, Table S3] The STP logic assumes that the three training algorithms differ only in whether temporal gradients and hidden state are propagated. This is not the case: SDBP (Eq. 5) removes the backward temporal term while retaining forward dynamics, which changes the optimization landscape and gradient variance; NoTD (Eq. 6) removes the leaky state entirely, changing model capacity. More fundamentally, the readout for several tasks averages outputs over time steps, so NoTD can still integrate temporal information at the readout level even without hidden state; the probe therefore does not isolate 'temporal information content' in the sense claimed by rules 1-3. The paper's own Table S3 shows the confound is real: SDBP exceeds STBP on CIFAR10-DVS (+0.50) and DVS-Gesture (+0.69), and NoTD exceeds STBP on CIFAR10-DVS (+1.50). Because the central claim that NSA is more effective than existing benchmarks depends entirely on the STP comparison, additional controls (e.g., non-temporal versions of the same tasks, multiple model families, repeated seeds) are needed before this conclusion can be accepted.
  2. [Sec. 3.1, Tables 2-4] All results are from single runs with no repeated-seed statistics or significance tests. Some of the gaps used to justify benchmark effectiveness are modest (e.g., AD: 1.29 dB difference between STBP and SDBP; AL: 5.00 percentage points), and without variance estimates it is impossible to determine whether these gaps are reliable or whether the model rankings in Tables 3-4 are robust. The authors should report means and standard deviations over at least three independent seeds, or otherwise justify the reproducibility of the reported numbers.
  3. [Sec. 3.3, Table 5] The energy-efficiency measurements are computed from spike statistics collected on a single task (AL) with a single sequence length (400) and then discussed as general properties of the models (e.g., 'LTC presents two orders of magnitude higher energy consumption'). Spike rates and network activity are task-dependent, so these numbers cannot be read as benchmark-wide efficiency measures. The authors should either report energy per task for the main comparisons or clearly restrict the energy conclusions to the AL setting.
  4. [Sec. 2.2 (AL) and Sec. 1] The abstract and introduction state that NSA includes 'seven real-world temporal processing tasks,' but AL is explicitly introduced as a synthetic dataset with the temporal dependency built in by construction (the final x-coordinate is a function of the action sequence). Using STP on AL to demonstrate that NSA tasks 'contain essential temporal dependencies' is therefore partially self-referential. To support the benchmark's validity, the authors should either validate on independently selected tasks or present an a-priori justification for the task-selection criterion.
  5. [Sec. 3.1, Tables S2-S4] The conclusion that existing benchmarks are inadequate is drawn from a single model family (LIF-SFNN) and a single training setup (STBP). Since the STP behavior may depend on the model class (e.g., recurrent or state-space architectures), the authors should verify the diagnostic on at least one additional architecture before concluding that the existing benchmarks 'fail to capture rich temporal dynamics' generally.
minor comments (6)
  1. [Abstract] The phrase 'seven real-world temporal processing tasks' is inaccurate because AL is a synthetic dataset; please revise to 'six real-world and one synthetic task' or list the task origins.
  2. [Table 3] The '*' for CE-LIF (SRNN) in the AD column is unexplained in the caption; please add a footnote explaining why the model is not applicable.
  3. [Table 5] The row for 'Spiking TCN' does not specify a neuron model; if LIF is used, please state this in the table or caption.
  4. [Sec. 4] The reference [Tay et al., 2020] is to the Long Range Arena benchmark for Transformers, not to an SNN language-modeling effort; the sentence 'preliminary efforts have been made to apply SNNs to long-term language modeling tasks' is not supported by this citation.
  5. [Sec. 2.3] Rules 1-3 lack a quantitative operationalization of 'comparable' and 'significant'; please define a threshold (e.g., relative performance drop) before applying the probe.
  6. [Table 2] The AD differences (1.29 dB and 2.00 dB) are described as 'substantial,' but no effect-size criterion is given for SI-SNR; please justify why these differences are meaningful.

Circularity Check

2 steps flagged · score 6.0 of 10

STP validation of NSA is partially self-definitional, and the AL task's sequence length was explicitly tuned to highlight the STP gap, making the benchmark-effectiveness claim for AL circular.

  1. self definitional [Sec. 2.3 (STP rules) and Sec. 3.1 (Benchmark Effectiveness Validation)]
    "In contrast, significant performance degradation in SDBP and NoTD compared to STBP indicates that the task contains rich temporal information and can effectively evaluate temporal processing capacity. ... As shown in Table 2, both SDBP and NoTD exhibit substantial performance degradation compared to STBP across all tasks in NSA. This suggests that the seven selected tasks contain essential temporal dependencies that must be effectively captured to attain high performance. Consequently, the proposed NSA serves as a more effective benchmark for neuromorphic temporal processing."

    Rule 3 of STP defines 'contains rich temporal information' operationally as the presence of a large SDBP/NoTD performance gap relative to STBP. Section 3.1 then observes exactly such gaps in Table 2 and uses them to conclude that the NSA tasks contain essential temporal dependencies and that NSA is a more effective benchmark. This is the criterion being restated as its own evidence: the conclusion follows directly from the operational definition, with no independent validation that STP performance gaps measure temporal content rather than optimization side effects. The paper's own Table S3, where SDBP and NoTD beat STBP on event-based datasets, further shows that such gaps are not uniquely attributable to temporal dependencies.

  2. fitted input called prediction [Supplementary Materials, 'Task-Specific Configurations', first paragraph]
    "In Benchmark Effectiveness Validation (Sec. 3.1), the sequence length of the AL task is adjusted to 200 thereby highlighting the performance gap among the three training algorithms in STP."

    This sentence explicitly states that the AL configuration used in the STP validation was adjusted in order to highlight the performance gap. The Table 2 AL result is therefore not an independent confirmation that the task contains rich temporal dependencies; it is a consequence of the chosen hyperparameter. The paper then presents this tuned gap as evidence that AL and NSA are effective temporal-processing benchmarks. For the AL task, the validation outcome is forced by construction: the benchmark parameter was fitted to the outcome later reported as validation.

full rationale

The NSA benchmark itself contains substantial independent value: Tables 3-5 report extensive comparisons of spiking neuron models, architectures, training speed, memory, and energy efficiency that are self-contained empirical measurements and do not depend on STP. No load-bearing self-citation chain was found, and STBP is an externally established algorithm. However, the paper's central claim that NSA is 'more effective' than existing benchmarks rests on the STP validation in Section 3.1, and that validation is partially circular in two concrete ways. First, the STP rule equates 'rich temporal information' with the existence of SDBP/NoTD performance gaps, so observing those gaps in Table 2 and then asserting temporal richness is a restatement of the operational definition rather than an independent test. Second, the supplementary materials admit that the AL sequence length in the STP validation was adjusted to highlight the performance gap, meaning the AL evidence was fitted to the desired outcome and then used as confirmation. Table S3 reinforces the concern by showing that SDBP and NoTD can outperform STBP on event-based datasets that STP classifies as lacking temporal dependencies, indicating that algorithmic gaps do not isolate temporal content. These issues compromise the 'effectiveness validation' argument, but they do not invalidate the benchmark's use for comparing SNN approaches. Score 6 reflects partial circularity confined to the validation claim.

Assumptions & free parameters 3 free parameters · 5 assumptions · 1 invented entities

The central claims rest on the STP ablation logic, the LIF reference model, and energy constants from the literature. No parameters are fitted to produce a phenomenon, but several design choices (task sequence lengths, action distributions) are selected by hand and not subject to sensitivity analysis.

free parameters (3)
  • AL action distribution probabilities = [0.05, 0.05, 0.45, 0.45]
    Chosen by hand to define the synthetic AL task; temporal dependency is deliberately embedded.
  • Sequence lengths for tasks = AL 200/400, EEG-MI 500, ALR 200, SSL 500, HAR 200
    Chosen to balance training cost and temporal richness; results could vary with different lengths.
  • Model-specific hyperparameters (hidden dimensions, learning rates, etc.) = Per task in Tables S5-S6
    Selected by hand per task to match parameter counts; no hyperparameter sensitivity analysis is reported.
assumptions (5)
  • domain assumption LIF neuron dynamics (Eq. 1-2) are a valid model for studying temporal processing
    The STP diagnostic and all baseline experiments rely on the LIF neuron model as the reference for temporal dynamics.
  • standard math Backpropagation through time (STBP, Eq. 3-4) correctly computes temporal gradients
    The STP comparison assumes STBP gradients are the correct reference for temporal learning.
  • domain assumption Removing backward temporal propagation (SDBP) and all temporal propagation (NoTD) isolates the contribution of temporal dependencies
    This is the load-bearing assumption of the STP tool; optimization side effects are assumed negligible.
  • domain assumption Energy constants EAC=0.9pJ and EMAC=4.6pJ from a 45nm CMOS process apply to the evaluated models
    Used to compute empirical energy cost in Table 5; actual hardware may differ.
  • domain assumption Training speed and memory measurements on an RTX 3090 are representative of general GPU training
    Used to draw conclusions about model efficiency in Sec. 3.3.
invented entities (1)
  • AL synthetic dataset
    purpose: Provides a controllable long-term temporal dependency task for the benchmark
    Generated by the authors from hand-specified action rules; no external validation or comparison against real robot data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Neuromorphic Sequential Arena: A Benchmark for Neuromorphic Temporal Processing." pith.science (2026). https://pith.science/paper/BXU442WT

@misc{pith2026250522035,
  author       = {Pith},
  title        = {Pith review of: Neuromorphic Sequential Arena: A Benchmark for Neuromorphic Temporal Processing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BXU442WT}},
  note         = {Machine review of arXiv:2505.22035}
}
read the original abstract

Temporal processing is vital for extracting meaningful information from time-varying signals. Recent advancements in Spiking Neural Networks (SNNs) have shown immense promise in efficiently processing these signals. However, progress in this field has been impeded by the lack of effective and standardized benchmarks, which complicates the consistent measurement of technological advancements and limits the practical applicability of SNNs. To bridge this gap, we introduce the Neuromorphic Sequential Arena (NSA), a comprehensive benchmark that offers an effective, versatile, and application-oriented evaluation framework for neuromorphic temporal processing. The NSA includes seven real-world temporal processing tasks from a diverse range of application scenarios, each capturing rich temporal dynamics across multiple timescales. Utilizing NSA, we conduct extensive comparisons of recently introduced spiking neuron models and neural architectures, presenting comprehensive baselines in terms of task performance, training speed, memory usage, and energy efficiency. Our findings emphasize an urgent need for efficient SNN designs that can consistently deliver high performance across tasks with varying temporal complexities while maintaining low computational costs. NSA enables systematic tracking of advancements in neuromorphic algorithm research and paves the way for developing effective and efficient neuromorphic temporal processing systems.

Figures

Figures reproduced from arXiv: 2505.22035 by the authors.

Figure 1
Figure 1. Comparison of the three training algorithms in STP. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 39 canonical work pages

  1. [1]

    A low power, fully event- based gesture recognition system

    [Amir et al., 2017] Arnon Amir, Brian Taba, David Berg, Timothy Melano, Jeffrey McKinstry, Carmelo Di Nolfo, Tapan Nayak, Alexander Andreopoulos, Guillaume Gar- reau, Marcela Mendoza, et al. A low power, fully event- based gesture recognition system. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recog- nition, pages 7243–7252,

    work page 2017

  2. [4]

    Aishell-1: An open-source mandarin speech corpus and a speech recognition baseline

    [Bu et al., 2017] Hui Bu, Jiayu Du, Xingyu Na, Bengu Wu, and Hao Zheng. Aishell-1: An open-source mandarin speech corpus and a speech recognition baseline. In 2017 20th conference of the oriental chapter of the international coordinating committee on speech databases and speech I/O systems and assessment (O-COCOSDA) , pages 1–5. IEEE,

    work page 2017

  3. [5]

    A review of the integrate-and-fire neuron model: I

    [Burkitt, 2006] Anthony N Burkitt. A review of the integrate-and-fire neuron model: I. homogeneous synap- tic input. Biological cybernetics, 95:1–19,

    work page 2006

  4. [10]

    Temporal efficient training of spik- ing neural network via gradient re-weighting

    [Deng et al., 2022] Shikuang Deng, Yuhang Li, Shanghang Zhang, and Shi Gu. Temporal efficient training of spik- ing neural network via gradient re-weighting. In Inter- national Conference on Learning Representations , pages 1–14,

    work page 2022

  5. [12]

    Towards Ultra-Low-Power Neuromorphic Speech Enhancement with Spiking-FullSubNet

    [Hao et al., 2024] Xiang Hao, Chenxiang Ma, Qu Yang, Jibin Wu, and Kay Chen Tan. Towards ultra-low- power neuromorphic speech enhancement with spiking- fullsubnet. arXiv preprint arXiv:2410.04785,

    work page Pith review arXiv 2024

  6. [13]

    Network model with internal complexity bridges artificial intelligence and neu- roscience

    [He et al., 2024] Linxuan He, Yunhui Xu, Weihua He, Yihan Lin, Yang Tian, Yujie Wu, Wenhui Wang, Ziyang Zhang, Junwei Han, Yonghong Tian, et al. Network model with internal complexity bridges artificial intelligence and neu- roscience. Nature Computational Science, 4(8):584–599,

    work page 2024

  7. [14]

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

    [Horowitz, 2014] Mark Horowitz. 1.1 computing’s energy problem (and what we can do about it). In 2014 IEEE International Solid-State Circuits Conference Digest of Technical Papers (ISSCC), pages 10–14. IEEE,

    work page 2014

  8. [16]

    Learning multiple layers of features from tiny images

    [Krizhevsky and Hinton, 2009] Alex Krizhevsky and Geof- frey Hinton. Learning multiple layers of features from tiny images. Technical Report 0, University of Toronto, Toronto, Ontario,

    work page 2009

Show all 44 references
  1. [20]

    Darwin3: a large-scale neu- romorphic chip with a novel isa and on-chip learning

    [Ma et al., 2024] De Ma, Xiaofei Jin, Shichun Sun, Yitao Li, Xundong Wu, Youneng Hu, Fangchao Yang, Huajin Tang, Xiaolei Zhu, Peng Lin, et al. Darwin3: a large-scale neu- romorphic chip with a novel isa and on-chip learning. Na- tional Science Review, 11(5):nwae102,

    2024

  2. [21]

    Networks of spiking neu- rons: the third generation of neural network models

    [Maass, 1997] Wolfgang Maass. Networks of spiking neu- rons: the third generation of neural network models. Neu- ral Networks, 10(9):1659–1671,

    1997

  3. [24]

    Sloclas: A database for joint sound localization and classification

    [Qian et al., 2021] Xinyuan Qian, Bidisha Sharma, Amine El Abridi, and Haizhou Li. Sloclas: A database for joint sound localization and classification. arXiv preprint arXiv:2108.02539,

    2021 arXiv

  4. [26]

    Delayed memory unit: Modeling temporal dependency through delay gate

    [Sun et al., 2024] Pengfei Sun, Jibin Wu, Malu Zhang, Paul Devos, and Dick Botteldooren. Delayed memory unit: Modeling temporal dependency through delay gate. IEEE Transactions on Neural Networks and Learning Systems ,

    2024

  5. [27]

    Multi-grained spatio- temporal features perceived network for event-based lip- reading

    [Tan et al., 2022] Ganchao Tan, Yang Wang, Han Han, Yang Cao, Feng Wu, and Zheng-Jun Zha. Multi-grained spatio- temporal features perceived network for event-based lip- reading. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 20094– 20103,

    2022

  6. [28]

    Long range arena: A benchmark for efficient transformers

    [Tay et al., 2020] Yi Tay, Mostafa Dehghani, Samira Abnar, Yikang Shen, Dara Bahri, Philip Pham, Jinfeng Rao, Liu Yang, Sebastian Ruder, and Donald Metzler. Long range arena: A benchmark for efficient transformers. arXiv preprint arXiv:2011.04006,

    2020 arXiv

  7. [29]

    The intel neuromorphic dns challenge

    [Timcheck et al., 2023] Jonathan Timcheck, Sumit Bam Shrestha, Daniel Ben Dayan Rubin, Adam Kupryjanow, Garrick Orchard, Lukasz Pindor, Timothy Shea, and Mike Davies. The intel neuromorphic dns challenge. Neuromor- phic Computing and Engineering, 3(3):034005,

    2023

  8. [30]

    Event-stream representation for human gaits identification using deep neural networks

    [Wang et al., 2022] Yanxiang Wang, Xian Zhang, Yiran Shen, Bowen Du, Guangrong Zhao, Lizhen Cui, and Hongkai Wen. Event-stream representation for human gaits identification using deep neural networks. IEEE Transactions on Pattern Analysis and Machine Intelli- gence, 44(7):3436–3449,

    2022

  9. [31]

    Hardvs: Revisiting human activity recog- nition with dynamic vision sensors

    [Wang et al., 2024] Xiao Wang, Zongzhen Wu, Bo Jiang, Zhimin Bao, Lin Zhu, Guoqi Li, Yaowei Wang, and Yonghong Tian. Hardvs: Revisiting human activity recog- nition with dynamic vision sensors. Proceedings of the AAAI Conference on Artificial Intelligence , 38(6):5615– 5623, Mar

    2024

  10. [32]

    Speech commands: A dataset for limited-vocabulary speech recognition

    [Warden, 2018] Pete Warden. Speech commands: A dataset for limited-vocabulary speech recognition. arXiv preprint arXiv:1804.03209,

    2018 arXiv

  11. [33]

    Espnet: End-to-end speech processing toolkit

    [Watanabe et al., 2018] Shinji Watanabe, Takaaki Hori, Shigeki Karita, Tomoki Hayashi, Jiro Nishitoba, Yuya Unno, Nelson Enrique Yalta Soplin, Jahn Heymann, Matthew Wiesner, Nanxin Chen, et al. Espnet: End-to-end speech processing toolkit. arXiv preprint arXiv:1804.00015,

    2018 arXiv

  12. [34]

    Wisdm smartphone and smartwatch activity and biometrics dataset

    [Weiss, 2019] Gary M Weiss. Wisdm smartphone and smartwatch activity and biometrics dataset. UCI Ma- chine Learning Repository: WISDM Smartphone and Smartwatch Activity and Biometrics Dataset Data Set , 7:133190–133202,

    2019

  13. [35]

    Spatio-temporal backpropagation for training high-performance spiking neural networks

    [Wu et al., 2018] Yujie Wu, Lei Deng, Guoqi Li, Jun Zhu, and Luping Shi. Spatio-temporal backpropagation for training high-performance spiking neural networks. Fron- tiers in Neuroscience, 12:331,

    2018

  14. [36]

    A vision chip with complementary pathways for open-world sensing

    [Yang et al., 2024] Zheyu Yang, Taoyi Wang, Yihan Lin, Yuguo Chen, Hui Zeng, Jing Pei, Jiazheng Wang, Xue Liu, Yichun Zhou, Jianqiang Zhang, et al. A vision chip with complementary pathways for open-world sensing. Nature, 629(8014):1027–1033,

    2024

  15. [37]

    Spike- driven transformer

    [Yao et al., 2024] Man Yao, Jiakui Hu, Zhaokun Zhou, Li Yuan, Yonghong Tian, Bo Xu, and Guoqi Li. Spike- driven transformer. Advances in neural information pro- cessing systems, 36,

    2024

  16. [38]

    Accurate and efficient time-domain classification with adaptive spiking recurrent neural networks

    [Yin et al., 2021] Bojian Yin, Federico Corradi, and Sander M Boht ´e. Accurate and efficient time-domain classification with adaptive spiking recurrent neural networks. Nature Machine Intelligence , 3(10):905–913,

    2021

  17. [39]

    Accurate online training of dynamical spiking neural networks through forward propagation through time

    [Yin et al., 2023] Bojian Yin, Federico Corradi, and Sander M Boht ´e. Accurate online training of dynamical spiking neural networks through forward propagation through time. Nature Machine Intelligence, 5(5):518–527,

    2023

  18. [40]

    Tc-lif: A two-compartment spiking neuron model for long-term sequential modelling

    [Zhang et al., 2024] Shimin Zhang, Qu Yang, Chenxiang Ma, Jibin Wu, Haizhou Li, and Kay Chen Tan. Tc-lif: A two-compartment spiking neuron model for long-term sequential modelling. In Proceedings of the AAAI Confer- ence on Artificial Intelligence , volume 38, pages 16838– 16847,

    2024

  19. [41]

    Temporal dendritic heterogeneity incorporated with spiking neural networks for learning multi-timescale dynamics

    [Zheng et al., 2024] Hanle Zheng, Zhong Zheng, Rui Hu, Bo Xiao, Yujie Wu, Fangwen Yu, Xue Liu, Guoqi Li, and Lei Deng. Temporal dendritic heterogeneity incorporated with spiking neural networks for learning multi-timescale dynamics. Nature Communications, 15(1):277,

    2024

  20. [42]

    En- hancing snn-based spatio-temporal learning: A benchmark dataset and cross-modality attention model

    [Zhou et al., 2024] Shibo Zhou, Bo Yang, Mengwen Yuan, Runhao Jiang, Rui Yan, Gang Pan, and Huajin Tang. En- hancing snn-based spatio-temporal learning: A benchmark dataset and cross-modality attention model. Neural Net- works, 180:106677,

    2024

  21. [400]

    left”, otherwise “right

    All networks are configured with hidden dimensions of [256, 256], and the second hidden layer is consistently utilized for compar- isons across evaluated models. During inference, we record the average spike frequency per time step, sample, and neuron for all spiking neuron po...

    2014

  22. [500]

    Notably, raw time-domain signals are processed directly, without the application of any frequency spectral analysis

    To simulate real-world application scenarios, directional background noise from 0°, 90°, 180°, and 270° provided in the dataset are randomly segmented and added to each sample at 0 SNR. Notably, raw time-domain signals are processed directly, without the application of any fre...

    2022

  23. [1997]

    Converting static image datasets to spiking neuromorphic datasets using sac- cades

    [Orchard et al., 2015] Garrick Orchard, Ajinkya Jayawant, Gregory K Cohen, and Nitish Thakor. Converting static image datasets to spiking neuromorphic datasets using sac- cades. Frontiers in neuroscience, 9:437,

    2015

  24. [1998]

    Eeg dataset and openbmi toolbox for three bci paradigms: an investigation into bci illiteracy

    [Lee et al., 2019] Min-Ho Lee, O-Yeon Kwon, Yong- Jeong Kim, Hong-Kyung Kim, Young-Eun Lee, John Williamson, Siamac Fazli, and Seong-Whan Lee. Eeg dataset and openbmi toolbox for three bci paradigms: an investigation into bci illiteracy. GigaScience, 8(5):giz002, 01

    2019

  25. [2006]

    Unleashing the potential of spiking neural networks for sequential modeling with contextual embedding

    [Chen et al., 2023] Xinyi Chen, Jibin Wu, Huajin Tang, Qinyuan Ren, and Kay Chen Tan. Unleashing the potential of spiking neural networks for sequential modeling with contextual embedding. arXiv preprint arXiv:2308.15150,

    2023 arXiv

  26. [2009]

    Gradient-based learning ap- plied to document recognition

    [LeCun et al., 1998] Yann LeCun, L ´eon Bottou, Yoshua Bengio, and Patrick Haffner. Gradient-based learning ap- plied to document recognition. Proceedings of the IEEE, 86(11):2278–2324,

    1998

  27. [2014]

    A review on machine learning for eeg signal processing in bioengineering

    [Hosseini et al., 2020] Mohammad-Parsa Hosseini, Amin Hosseini, and Kiarash Ahi. A review on machine learning for eeg signal processing in bioengineering. IEEE reviews in biomedical engineering, 14:204–218,

    2020

  28. [2015]

    Towards artificial general intelligence with hybrid tianjic chip architecture

    [Pei et al., 2019] Jing Pei, Lei Deng, Sen Song, Mingguo Zhao, Youhui Zhang, Shuang Wu, Guanrui Wang, Zhe Zou, Zhenzhi Wu, Wei He, et al. Towards artificial general intelligence with hybrid tianjic chip architecture. Nature, 572(7767):106–111,

    2019

  29. [2017]

    An empirical evaluation of generic convolutional and recurrent networks for sequence modeling

    [Bai et al., 2018] Shaojie Bai, J Zico Kolter, and Vladlen Koltun. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. arXiv preprint arXiv:1803.01271,

    2018 arXiv

  30. [2018]

    Long short-term memory and learning-to-learn in net- works of spiking neurons

    [Bellec et al., 2018] Guillaume Bellec, Darjan Salaj, Anand Subramoney, Robert Legenstein, and Wolfgang Maass. Long short-term memory and learning-to-learn in net- works of spiking neurons. Advances in neural information processing systems, 31,

    2018

  31. [2019]

    Cifar10-dvs: an event-stream dataset for object classification.Frontiers in Neuroscience, 11:244131,

    [Li et al., 2017] Hongmin Li, Hanchao Liu, Xiangyang Ji, Guoqi Li, and Luping Shi. Cifar10-dvs: an event-stream dataset for object classification.Frontiers in Neuroscience, 11:244131,

    2017

  32. [2020]

    Loihi: A neuromorphic manycore processor with on-chip learning

    [Davies et al., 2018] Mike Davies, Narayan Srinivasa, Tsung-Han Lin, Gautham Chinya, Yongqiang Cao, Sri Harsha Choday, Georgios Dimou, Prasad Joshi, Nabil Imam, Shweta Jain, et al. Loihi: A neuromorphic manycore processor with on-chip learning. IEEE Micro, 38(1):82–99,

    2018

  33. [2021]

    Learning long sequences in spiking neural networks

    [Stan and Rhodes, 2024] Matei-Ioan Stan and Oliver Rhodes. Learning long sequences in spiking neural networks. Scientific Reports, 14(1):21957,

    2024

  34. [2022]

    Parallel spiking neurons with high efficiency and ability to learn long-term dependen- cies

    [Fang et al., 2023] Wei Fang, Zhaofei Yu, Zhaokun Zhou, Ding Chen, Yanqi Chen, Zhengyu Ma, Timoth´ee Masque- lier, and Yonghong Tian. Parallel spiking neurons with high efficiency and ability to learn long-term dependen- cies. In Advances in Neural Information Processing Sys- ...

    2023

  35. [2023]

    Pmsn: A parallel multi-compartment spiking neuron for multi-scale temporal processing

    [Chen et al., 2024] Xinyi Chen, Jibin Wu, Chenxiang Ma, Yinsong Yan, Yujie Wu, and Kay Chen Tan. Pmsn: A parallel multi-compartment spiking neuron for multi-scale temporal processing. arXiv preprint arXiv:2408.14917 ,

    2024 arXiv

  36. [2024]

    The heidel- berg spiking data sets for the systematic evaluation of spik- ing neural networks

    [Cramer et al., 2020] Benjamin Cramer, Yannik Stradmann, Johannes Schemmel, and Friedemann Zenke. The heidel- berg spiking data sets for the systematic evaluation of spik- ing neural networks. IEEE Transactions on Neural Net- works and Learning Systems, 33(7):2744–2757,

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.