Pith. sign in

REVIEW 4 major objections 2 minor 2 cited by

Universal Learning of Nonlinear Dynamics

T0 review · 4 major / 2 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 This paper proves that a spectral filtering algorithm achieves vanishing prediction error for any nonlinear dynamical system with finitely many marginally stable modes, with rates governed by a new quantitative control-theoretic notion of…

desk verdict The abstract promises a universal learning guarantee for nonlinear dynamics, but the supplied full text is a different micro-expression paper, so this submission is not reviewable. read the letter →

arxiv 2508.11990 v1 pith:JNFHQWM6 submitted 2025-08-16 cs.LG math.OCstat.ML

Evan Dogariu , Anand Brahmbhatt , Elad Hazan This is my paper · ORCID
classification cs.LGmath.OCstat.ML
keywords nonlineardynamicalsystemsspectralfilteringlearningfromobservationsmarginallystablemodesonlineconvexoptimizationpredictionerrorlearnabilitycontroltheory
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

This paper asks whether an unknown nonlinear dynamical system can be learned from observations alone, without a model. The authors claim a universal guarantee: a spectral filtering algorithm, run on a window of past observations, predicts the next observation with error that vanishes in the limit, for any nonlinear system having finitely many marginally stable modes. The rate at which the error vanishes is governed by a newly introduced quantitative control-theoretic measure of learnability. The supporting technical contribution is a generalized spectral filtering algorithm for linear systems that handles asymmetric dynamics and noise correction. If the proof is right, this would be the first universal learning guarantee for this broad class of systems.

What carries the argument

The central object is the spectral filtering algorithm, a technique that represents a dynamical system through a spectral decomposition of the matrix built from past observations, then uses online convex optimization to update predictions. The new version extends the original spectral filtering method to asymmetric linear dynamics and adds a noise-correction component, which is what allows the method to cover nonlinear systems that are well described by finitely many marginally stable modes. This spectral representation is the mechanism that lets the algorithm capture the modes that dominate the system's behaviour without requiring a known state-space model.

What would settle it

Run the spectral filtering algorithm on a nonlinear system with finitely many marginally stable modes, such as a Duffing oscillator or a coupled oscillator network, and check whether the prediction error actually decreases to zero as the number of observations grows; a plateau above zero would refute the vanishing-error claim. Alternatively, compute the proposed learnability measure for a specific system: if the measure is not a finite, well-defined quantity for a system that the algorithm still predicts well, the rate statement would need revision.

Watch

Extended reading notes

Core claim

The paper's central claim is that vanishing prediction error holds for every nonlinear dynamical system with finitely many marginally stable modes when learning is done by spectral filtering. The algorithm learns a mapping from past observations to the next observation based on a spectral representation of the system, and the analysis is carried out with tools from online convex optimization. The error rate is expressed through a new quantitative control-theoretic notion of learnability, so the guarantee is not just qualitative but carries a rate. The main technical step is a new spectral filtering algorithm for linear dynamical systems that incorporates past observations and applies to general noisy and marginally stable systems, including asymmetric dynamics; this is presented as a result of independent interest.

Load-bearing premise

The load-bearing premise is that a nonlinear system with finitely many marginally stable modes is adequately represented by the spectral decomposition the algorithm uses; the abstract offers no construction showing how those modes are identified from data or why the spectral representation captures them.

Editorial extensions

If this is right

  • Any nonlinear system with finitely many marginally stable modes can be predicted online, with the prediction error shrinking to zero as more observations are collected.
  • The generalized linear spectral filter extends spectral filtering to asymmetric and noisy linear dynamics, a regime the original method did not cover.
  • Error rates are expressed through a single quantitative learnability measure, so systems that score higher on this measure are predicted with faster convergence.
  • Because the algorithm is online, it can be deployed in streaming settings without a separate training phase or a known state-space model.

Reading between the lines

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

  • If the guarantee is correct, empirical estimation of the learnability measure from a single trajectory could let practitioners predict in advance how quickly prediction error will fall; the paper does not propose such an estimator.
  • The noise-correction component suggests potential use in closed-loop control of marginally stable systems, though the abstract does not address control or closed-loop stability.
  • The finite-mode assumption is the exact boundary of the claim: systems with infinitely many marginally stable modes fall outside the guarantee, even if they are otherwise similar.
  • Spatially extended systems such as fluid flows, where low-dimensional marginally stable modes dominate, would form a natural test bed if the finite-mode condition holds; this extension is not in the paper.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 2 minor

Summary. The manuscript arXiv:2508.11990 is announced as a learning-theory paper: it claims a universal learning guarantee for nonlinear dynamical systems that have finitely many marginally stable modes, achieved by a spectral filtering algorithm with vanishing prediction error and rates governed by a novel quantitative control-theoretic learnability measure. It also claims a new spectral filtering method for linear systems that applies to asymmetric dynamics and includes noise correction. However, the supplied full text is not this paper: it is an unrelated computer-vision paper on event-based facial micro-expression analysis, with an SNN baseline for Action Unit recognition and a conditional variational autoencoder for frame reconstruction. As a result, none of the announced theorem statements, definitions, proofs, or algorithmic details are present, and the central claim cannot be inspected or verified from the submitted material.

Significance. If the claimed result were true, it would be a substantial contribution: a first universal learning guarantee for a broad class of nonlinear dynamical systems, together with an extension of spectral filtering to asymmetric and noisy linear systems. Such a result would likely be of interest to the online learning and system identification communities. That said, the submitted manuscript provides no evidence for the claim. There are no theorem statements, no definitions of the system class or of the learnability measure, no proofs, no algorithm descriptions, and no reproducible code or derivations. The only actual content in the full text is a preliminary dataset paper on micro-expression analysis, which does not bear on the announced learning-theory result. The significance of the claimed contribution cannot be assessed because the contribution itself is absent from the manuscript.

major comments (4)
  1. [Full text (pp. 1-10)] The body of the submission is a different paper, on event-based facial micro-expression analysis, not the learning-theory manuscript promised in the abstract. No theorem, proof, or spectral filtering algorithm appears anywhere in the text. The central claim of the abstract is therefore unsupported by any of the supplied content.
  2. [Abstract (opening sentence)] The phrase 'finitely many marginally stable modes' is undefined for nonlinear dynamical systems. The manuscript gives no state-space dimension, observation model, notion of mode for nonlinear systems, or construction of the spectral representation claimed to capture these modes. Without these definitions, the universal guarantee cannot be formulated, let alone verified.
  3. [Abstract (second sentence)] The 'novel quantitative control-theoretic notion of learnability' is neither defined nor shown to be computable. Because the rates are asserted to be governed by this measure, the claim can only be evaluated if the measure is given explicitly; otherwise there is no way to rule out that the measure was chosen to absorb the error terms.
  4. [Abstract (final sentences)] The claimed generalization of spectral filtering to asymmetric dynamics and noise correction is not described. No algorithm, update rule, or comparison with the original spectral filtering algorithm is provided, so the 'independent interest' claim cannot be checked.
minor comments (2)
  1. [Full text, first page] The header of the full text identifies it as arXiv:2508.11988v2, while the submitted abstract is for arXiv:2508.11990; this appears to be a submission mix-up that must be corrected.
  2. [Full text, Tables 3 and 4] The hardware specification tables contain incomplete column rendering; if the intended manuscript were the micro-expression paper, these tables would need reformatting. This is noted only for completeness, since the main issue is the paper mismatch.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity can be established: the claimed derivation is entirely in the abstract, and the supplied full text is an unrelated paper, so no input-to-output reduction is quotable.

full rationale

The circularity pass requires quoting the paper and exhibiting a specific reduction in which a claimed prediction or derivation is equivalent to its own inputs by construction. No such reduction is available here. The target manuscript arXiv:2508.11990 supplies only an abstract asserting a vanishing-prediction-error guarantee for marginally stable nonlinear systems, with rates governed by a new 'quantitative control-theoretic notion of learnability.' The supplied full text is arXiv:2508.11988v2, an unrelated computer vision paper on event-based facial micro-expression analysis, containing no definitions, theorem statements, proofs, or algorithm derivations matching the abstract. Because the body needed to verify whether the learnability measure, the mode representation, or the spectral filtering generalization is defined circularly is absent, the honest finding is not circularity but unverifiability. No self-citation chain, fitted parameter renamed as prediction, ansatz smuggled via citation, or uniqueness theorem imported from the authors can be identified from the available text. Under the hard rule that circularity may only be claimed when the specific reduction can be quoted, the score is 0.

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

The abstract alone cannot support an exhaustive ledger. No free parameters or invented entities are named. The listed axioms are the minimal domain assumptions implied by the abstract's claim. The 'novel quantitative control-theoretic notion of learnability' would be a key quantity if the body were present, but its definition is unavailable.

assumptions (3)
  • domain assumption The target system is marginally stable and has finitely many marginally stable modes.
    Stated in the abstract as the scope of the guarantee; the body does not define the class.
  • domain assumption Past observations provide sufficient information to estimate a spectral representation of the system.
    The abstract's learning claim presupposes identifiability from observations, which is not discussed in the abstract.
  • standard math Standard tools of online convex optimization and spectral theory apply as background.
    The abstract mentions online convex optimization as the proof technique; no derivation is shown.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Universal Learning of Nonlinear Dynamics." pith.science (2026). https://pith.science/paper/JNFHQWM6

@misc{pith2026250811990,
  author       = {Pith},
  title        = {Pith review of: Universal Learning of Nonlinear Dynamics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JNFHQWM6}},
  note         = {Machine review of arXiv:2508.11990}
}
read the original abstract

We study the fundamental problem of learning a marginally stable unknown nonlinear dynamical system. We describe an algorithm for this problem, based on the technique of spectral filtering, which learns a mapping from past observations to the next based on a spectral representation of the system. Using techniques from online convex optimization, we prove vanishing prediction error for any nonlinear dynamical system that has finitely many marginally stable modes, with rates governed by a novel quantitative control-theoretic notion of learnability. The main technical component of our method is a new spectral filtering algorithm for linear dynamical systems, which incorporates past observations and applies to general noisy and marginally stable systems. This significantly generalizes the original spectral filtering algorithm to both asymmetric dynamics as well as incorporating noise correction, and is of independent interest.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Spectral Distillation: From Nonlinear Dynamics to Linear State-Space Models

    cs.LG 2026-08 conditional novelty 6.0 of 10

    A convex spectral learner followed by spectral-to-LDS distillation extracts compact linear state-space predictors from nonlinear dynamics with a provable one-step error decomposition.

  2. SFO: Learning PDE Operators via Spectral Filtering

    cs.LG 2026-01 conditional novelty 6.0 of 10

    A neural operator that expands PDE kernels in fixed Hilbert-matrix eigenmodes achieves state-of-the-art benchmark accuracy with substantially fewer parameters.

Reference graph

Works this paper leans on

47 extracted references · 43 canonical work pages · cited by 2 Pith papers

  1. [1]

    Beyond rgb: Tri-modal microexpression recognition with rgb, thermal, and event data

    Mira Adra, N ´elida Mirabet-Herranz, and Jean-Luc Dugelay. Beyond rgb: Tri-modal microexpression recognition with rgb, thermal, and event data. In International Conference on Pattern Recognition, 2024. 2, 7

    work page 2024

  2. [2]

    Person re-identification without identification via event anonymiza- tion

    Shafiq Ahmad, Pietro Morerio, and Alessio Del Bue. Person re-identification without identification via event anonymiza- tion. 2023 IEEE/CVF International Conference on Com- puter Vision (ICCV), pages 11098–11107, 2023. 2

    work page 2023

  3. [3]

    Direct face detection and video reconstruction from event cameras

    Souptik Barua, Yoshitaka Miyatani, and Ashok Veeraragha- van. Direct face detection and video reconstruction from event cameras. In 2016 IEEE Winter Conference on Appli- cations of Computer Vision (WACV), pages 1–9, 2016. 3

    work page 2016

  4. [4]

    Federico Becattini, Lorenzo Berlincioni, Luca Cultrera, and A. Bimbo. Neuromorphic face analysis: a survey. Pattern Recognit. Lett., 187:42–48, 2024. 2

    work page 2024

  5. [5]

    Federico Becattini, Luca Cultrera, Lorenzo Berlincioni, Claudio Ferrari, Andrea Leonardo, and A. Bimbo. Neuro- morphic facial analysis with cross-modal supervision.ArXiv, abs/2409.10213, 2024. 2

    arXiv 2024

  6. [6]

    Lorenzo Berlincioni, Luca Cultrera, Chiara Albisani, Lisa Cresti, Andrea Leonardo, Sara Picchioni, Federico Becattini, and A. Bimbo. Neuromorphic event-based facial expression recognition. 2023 IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition Workshops (CVPRW) , pages 4109–4119, 2023. 2

    work page 2023

  7. [7]

    Lorenzo Berlincioni, Luca Cultrera, Federico Becattini, and A. Bimbo. Neuromorphic valence and arousal estimation. ArXiv, abs/2401.16058, 2024. 3

    arXiv 2024

  8. [8]

    Eddd: Event-based drowsiness driving detection through facial motion analysis with neuromorphic vision sensor

    Guang Chen, Lin Hong, Jinhu Dong, Peigen Liu, J ¨org Con- radt, and Alois Knoll. Eddd: Event-based drowsiness driving detection through facial motion analysis with neuromorphic vision sensor. IEEE Sensors Journal, 20:6170–6181, 2020. 1

    work page 2020

Show all 47 references
  1. [9]

    Spatio-temporal transformers for action unit classification with event cam- eras, 2024

    Luca Cultrera, Federico Becattini, Lorenzo Berlincioni, Claudio Ferrari, and Alberto Del Bimbo. Spatio-temporal transformers for action unit classification with event cam- eras, 2024. 7

    2024

  2. [10]

    Chinya, Yongqiang Cao, Sri Harsha Choday, Geor- gios D

    Mike Davies, Narayan Srinivasa, Tsung-Han Lin, Gau- tham N. Chinya, Yongqiang Cao, Sri Harsha Choday, Geor- gios D. Dimou, Prasad Joshi, Nabil Imam, Shweta Jain, Yuyun Liao, Chit-Kwan Lin, Andrew Lines, Ruokun Liu, Deepak A. Mathaikutty, Steve McCoy, Arnab Paul, Jonathan Tse, ...

    2018

  3. [11]

    Yiting Dong, Dongcheng Zhao, Yang Li, and Yi Zeng. An unsupervised stdp-based spiking neural network inspired by biologically plausible learning rules and connections.Neural networks : the official journal of the International Neural Network Society, 165:799–808, 2022. 3

    2022

  4. [12]

    Universal and cultural differences in facial ex- pression of emotion

    Paul Ekman. Universal and cultural differences in facial ex- pression of emotion. In Nebraska Symposium on Motivation, pages 207–283, 1971. 1

    1971

  5. [13]

    Paul Ekman and Wallace V . Friesen. Facial action coding system: a technique for the measurement of facial move- ment. 1978. 2, 4

    1978

  6. [14]

    Incorporating learnable membrane time constant to enhance learning of spiking neu- ral networks

    Wei Fang, Zhaofei Yu, Yanqing Chen, Timoth´ee Masquelier, Tiejun Huang, and Yonghong Tian. Incorporating learnable membrane time constant to enhance learning of spiking neu- ral networks. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 2641–2651, 2020. 3, 6

    2021

  7. [15]

    Friesen and Paul Ekman

    Wallace V . Friesen and Paul Ekman. Emfacs-7: Emotional facial action coding system. 1983. 3

    1983

  8. [16]

    v2e: From video frames to realistic dvs events

    Yuhuang Hu, Shih-Chii Liu, and Tobi Delbruck. v2e: From video frames to realistic dvs events. 2021 IEEE/CVF Con- ference on Computer Vision and Pattern Recognition Work- shops (CVPRW), pages 1312–1321, 2021. 2

    2021

  9. [17]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. CoRR, abs/1412.6980, 2014. 7

    2014 arXiv

  10. [18]

    Advances in facial expression recognition: A survey of methods, benchmarks, models, and datasets.Inf., 15:135, 2024

    Thomas Kopalidis, Vassilios Solachidis, Nicholas Vretos, and Petros Daras. Advances in facial expression recognition: A survey of methods, benchmarks, models, and datasets.Inf., 15:135, 2024. 1

    2024

  11. [19]

    Foster, Pamela Ventola, Fr ´ed´erick Shic, and Linda G

    Beibin Li, Sachin Mehta, Deepali Aneja, Claire E. Foster, Pamela Ventola, Fr ´ed´erick Shic, and Linda G. Shapiro. A facial affect analysis system for autism spectrum disorder. 2019 IEEE International Conference on Image Processing (ICIP), pages 4549–4553, 2019. 1

    2019

  12. [20]

    Deep learning for micro-expression recogni- tion: A survey

    Yante Li, Jinsheng Wei, Yang Liu, Janne Kauttonen, and Guoying Zhao. Deep learning for micro-expression recogni- tion: A survey. IEEE Transactions on Affective Computing, 13:2028–2046, 2021. 2

    2021

  13. [21]

    Sgdr: Stochastic gradient descent with warm restarts

    Ilya Loshchilov and Frank Hutter. Sgdr: Stochastic gradient descent with warm restarts. arXiv: Learning, 2016. 7

    2016

  14. [22]

    Cohn, Takeo Kanade, Jason M

    Patrick Lucey, Jeffrey F. Cohn, Takeo Kanade, Jason M. Saragih, Zara Ambadar, and I. Matthews. The extended cohn-kanade dataset (ck+): A complete dataset for action unit and emotion-specified expression.2010 IEEE Computer Society Conference on Computer Vision and Pattern Recog...

    2010

  15. [23]

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

    Wolfgang Maass. Networks of spiking neurons: The third generation of neural network models. Electron. Colloquium Comput. Complex., TR96, 1996. 2

    1996

  16. [24]

    Arthur, Rodrigo Alvarez-Icaza, An- drew S

    Paul Merolla, John V . Arthur, Rodrigo Alvarez-Icaza, An- drew S. Cassidy, Jun Sawada, Filipp Akopyan, Bryan L. Jackson, Nabil Imam, Chen Guo, Yutaka Nakamura, Bernard Brezzo, Ivan V o, Steven K. Esser, Rathinakumar Appuswamy, Brian Taba, Arnon Amir, Myron Flickner, William P....

    2014

  17. [25]

    Event-based asynchronous sparse con- volutional networks

    Nico Messikommer, Daniel Gehrig, Antonio Loquercio, and Davide Scaramuzza. Event-based asynchronous sparse con- volutional networks. ArXiv, abs/2003.09148, 2020. 2

    2003 arXiv

  18. [26]

    Neftci, Hesham Mostafa, and Friedemann Zenke

    Emre O. Neftci, Hesham Mostafa, and Friedemann Zenke. Surrogate gradient learning in spiking neural networks: Bringing the power of gradient-based optimization to spik- ing neural networks. IEEE Signal Processing Magazine, 36: 51–63, 2019. 2 9

    2019

  19. [27]

    High Frame Rate Video Re- construction Based on an Event Camera .IEEE Transactions on Pattern Analysis & Machine Intelligence , 44(05):2519– 2533, 2022

    Liyuan Pan, Richard Hartley, Cedric Scheerlinck, Miaomiao Liu, Xin Yu, and Yuchao Dai. High Frame Rate Video Re- construction Based on an Event Camera .IEEE Transactions on Pattern Analysis & Machine Intelligence , 44(05):2519– 2533, 2022. 3

    2022

  20. [28]

    Valstar, Ron Rademaker, and Ludo Maat

    Maja Pantic, Michel F. Valstar, Ron Rademaker, and Ludo Maat. Web-based database for facial expression analysis. 2005 IEEE International Conference on Multimedia and Expo, pages 5 pp.–, 2005. 2

    2005

  21. [29]

    A general psychoevolutionary theory of emotion

    Robert Plutchik. A general psychoevolutionary theory of emotion. 1980. 2

    1980

  22. [30]

    Perales L ´opez

    Silvia Ramis, Jose Maria Buades Rubio, and Francisco J. Perales L ´opez. Using a social robot to evaluate facial ex- pressions in the wild.Sensors (Basel, Switzerland), 20, 2020. 1

    2020

  23. [31]

    Esim: an open event camera simulator

    Henri Rebecq, Daniel Gehrig, and Davide Scaramuzza. Esim: an open event camera simulator. In Conference on Robot Learning, 2018. 2

    2018

  24. [32]

    Mahony, and Davide Scaramuzza

    Cedric Scheerlinck, Henri Rebecq, Daniel Gehrig, Nick Barnes, Robert E. Mahony, and Davide Scaramuzza. Fast image reconstruction with an event camera. In 2020 IEEE Winter Conference on Applications of Computer Vision (WACV), pages 156–163, 2020. 3

    2020

  25. [33]

    TenHouten

    Warren D. TenHouten. Basic emotion theory, social con- structionism, and the universal ethogram. Social Science In- formation, 60:610 – 630, 2021. 2

    2021

  26. [34]

    Learning spatiotemporal features with 3d convolutional networks

    Du Tran, Lubomir Bourdev, Rob Fergus, Lorenzo Torre- sani, and Manohar Paluri. Learning spatiotemporal features with 3d convolutional networks. In 2015 IEEE International Conference on Computer Vision (ICCV), pages 4489–4497,

    2015

  27. [35]

    Sl- animals-dvs: event-driven sign language animals dataset

    Ajay Vasudevan, Pablo Negri, Camila Di Ielsi, Bernab ´e Linares-Barranco, and Teresa Serrano-Gotarredona. Sl- animals-dvs: event-driven sign language animals dataset. Pattern Analysis and Applications, 25:505 – 520, 2021. 3

    2021

  28. [36]

    Event- based gesture and facial expression recognition: A compara- tive analysis

    Rodrigo Verschae and Ignacio Bugueno-Cordova. Event- based gesture and facial expression recognition: A compara- tive analysis. IEEE Access, 11:121269–121283, 2023. 2

    2023

  29. [37]

    evtrans- fer: A transfer learning framework for event-based facial ex- pression recognition, 2025

    Rodrigo Verschae and Ignacio Bugueno-Cordova. evtrans- fer: A transfer learning framework for event-based facial ex- pression recognition, 2025. 3

    2025

  30. [38]

    Real-time driver drowsiness detection using facial action units

    Malaika Vijay, Nandagopal Netrakanti Vinayak, Maanvi Nunna, and Natarajan Subramanyam. Real-time driver drowsiness detection using facial action units. 2020 25th International Conference on Pattern Recognition (ICPR) , pages 10113–10119, 2021. 1

    2020

  31. [39]

    Movellan

    Esra Vural, M ¨ujdat C ¸ etin, Ayt¨ul Erc ¸il, Gwen Littlewort, Mar- ian Stewart Bartlett, and Javier R. Movellan. Drowsy driver detection through facial movement analysis. In ICCV-HCI,

  32. [40]

    A multi-modal driver emotion dataset and study: In- cluding facial expressions and synchronized physiological signals

    Guoliang Xiang, Song Yao, Hanwen Deng, Xianhui Wu, Xinghua Wang, Qian Xu, Tianjian Yu, Kui Wang, and Yong Peng. A multi-modal driver emotion dataset and study: In- cluding facial expressions and synchronized physiological signals. Eng. Appl. Artif. Intell., 130:107772, 2024. 1

    2024

  33. [41]

    Joint face detection and facial expression recognition with mtcnn

    Jia Xiang and Gengming Zhu. Joint face detection and facial expression recognition with mtcnn. 2017 4th International Conference on Information Science and Control Engineering (ICISCE), pages 424–427, 2017. 5

    2017

  34. [42]

    Kashu Yamazaki, Viet-Khoa V o-Ho, D Bulsara, and Ngan T. H. Le. Spiking neural networks and their applications: A review. Brain Sciences, 12, 2022. 3

    2022

  35. [43]

    How fast are the leaked facial expressions: The duration of micro-expressions

    Wen-Jing Yan, Qi Wu, Jing Liang, Yu-Hsin Chen, and Xi- aolan Fu. How fast are the leaked facial expressions: The duration of micro-expressions. Journal of Nonverbal Behav- ior, 37:217–230, 2013. 2

    2013

  36. [44]

    Aide: A vision-driven multi-view, multi- modal, multi-tasking dataset for assistive driving perception

    Dingkang Yang, Shuai Huang, Zhi Xu, Zhenpeng Li, Shunli Wang, Mingcheng Li, Yuzheng Wang, Yang Liu, Kun Yang, Zhaoyu Chen, Yan Wang, Jing Liu, Pei Zhang, Peng Zhai, and Lihua Zhang. Aide: A vision-driven multi-view, multi- modal, multi-tasking dataset for assistive driving per...

    2023

  37. [45]

    Shan, and Xilin Chen

    Jiabei Zeng, S. Shan, and Xilin Chen. Facial expression recognition with inconsistently annotated datasets. In Eu- ropean Conference on Computer Vision, 2018. 2

    2018

  38. [46]

    V2ce: Video to continuous events simulator

    Zhongyang Zhang, Shuyang Cui, Kaidong Chai, Haowen Yu, Subhasish Dasgupta, Upal Mahbub, and Tauhidur Rah- man. V2ce: Video to continuous events simulator. 2024 IEEE International Conference on Robotics and Automation (ICRA), pages 12455–12461, 2023. 3

    2024

  39. [47]

    Learning to reconstruct high speed and high dynamic range videos from events

    Yunhao Zou, Yinqiang Zheng, Tsuyoshi Takatani, and Ying Fu. Learning to reconstruct high speed and high dynamic range videos from events. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2024–2033, 2021. 3 10

    2021

Pith tools

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