pith. sign in

arxiv: 2605.22139 · v1 · pith:AX5XF66Xnew · submitted 2026-05-21 · 💻 cs.CV

EventGait: Towards Robust Gait Recognition with Event Streams

Senyan Xu , Shuai Chen , Chuanfu Shen , Kean Liu , Zhijing Sun , Chengzhi Cao , Xueyang Fu This is my paper

Pith reviewed 2026-05-22 06:53 UTC · model grok-4.3

classification 💻 cs.CV
keywords gait recognitionevent camerasevent-based visionlow-light conditionsspiking neural networksdual-stream modelssynthetic datasets
0
0 comments X

The pith

Event-based gait recognition matches camera-based methods in normal light and outperforms them in low light by processing event streams with a dual-stream network.

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

The paper aims to show that gait recognition becomes more reliable in real-world conditions by replacing conventional cameras with event cameras, which record only brightness changes at microsecond resolution and maintain high dynamic range. This approach matters because standard gait systems degrade sharply under poor illumination or fast motion, restricting their use in security and monitoring. The authors present EventGait, an end-to-end model that keeps separate streams for dynamic motion and static shape to retain fine temporal details instead of collapsing events into static images. They also supply a synthesis pipeline that turns ordinary videos into event data and release two new event-based gait benchmarks. If the central claim holds, gait recognition can operate without special lighting or controlled environments.

Core claim

EventGait is an end-to-end dual-stream framework that separately models motion and shape while preserving the advantages of events. The dynamic stream uses a Mixture of Spiking Experts with diverse neuron constants for robust perception across complex scenes, and the static stream learns dense shape representations via Cross-modal Structure Alignment with large vision foundation models. This design achieves results comparable to camera-based gait recognition under normal conditions while significantly outperforming it in low-light scenarios and sets a new state of the art on both synthesized and real-world event-based gait benchmarks.

What carries the argument

Dual-stream architecture that processes raw event streams, with Mixture of Spiking Experts (MoSE) handling motion dynamics and Cross-modal Structure Alignment (CroSA) handling shape information.

If this is right

  • Gait recognition systems can maintain accuracy in dark or variable lighting without additional illumination.
  • Avoiding long-window event aggregation preserves the fine temporal cues needed to distinguish walking patterns.
  • The released SUSTech1K-E and CCGR-Mini-E benchmarks provide standardized test sets for comparing future event-based gait methods.
  • Joint training of dynamic and static streams allows the model to exploit complementary information present only in event data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same separation of motion and shape streams could transfer to other motion-based tasks such as action recognition under changing light.
  • Neuromorphic hardware implementations of the spiking experts may enable low-power, always-on identification systems.
  • Large-scale synthetic event datasets generated by the pipeline could accelerate progress on additional event-camera applications beyond biometrics.

Load-bearing premise

The synthesis pipeline that converts conventional gait videos into event streams produces data whose statistics and noise characteristics are sufficiently close to real event camera output for the reported performance gains to generalize.

What would settle it

Capture the same gait sequences with a real event camera in low-light conditions and measure whether EventGait still outperforms or matches camera-based accuracy on those direct recordings.

Figures

Figures reproduced from arXiv: 2605.22139 by Chengzhi Cao, Chuanfu Shen, Kean Liu, Senyan Xu, Shuai Chen, Xueyang Fu, Zhijing Sun.

Figure 1
Figure 1. Figure 1: Comparison of RGB and event cameras under day and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The workflow of our EventGait. The detailed architecture of the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: (a) The simplified schematic of the different spiking neurons’ dynamics across complex conditions ( [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The details of the Static Shape Stream, which is trained [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Cross-view performance comparison between Lidar [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
read the original abstract

Gait recognition enables non-intrusive, privacy-preserving identification but suffers in uncontrolled environments due to illumination and motion sensitivity of conventional cameras. In this work, we explore gait recognition using event cameras, which offer microsecond temporal resolution and high dynamic range, naturally capturing robust dynamic cues and suppressing static noise. Existing event-based approaches typically aggregate event streams into event images over long time windows, thereby discarding fine-grained motion dynamics critical for gait recognition. Therefore, we propose \textbf{EventGait}, an end-to-end dual-stream framework that separately models motion and shape while preserving the advantages of events. Our dynamic stream leverages a Mixture of Spiking Experts (MoSE) with diverse neuron constants for robust dynamic perception across complex motion and illumination scenes, while the static stream learns dense shape representations via Cross-modal Structure Alignment (CroSA) with large vision foundation models. To address the absence of large-scale event-based gait datasets, we introduce a synthesis pipeline and release two new benchmarks: SUSTech1K-E and CCGR-Mini-E. Extensive experiments have shown that event-based gait recognition not only achieves results comparable to camera-based gait recognition under normal conditions but also significantly outperforms it in low-light scenarios. Our approach sets a new state of the art on both synthesized and real-world event-based gait benchmarks, highlighting the robustness and potential of event-driven gait analysis. The code and datasets are released at https://github.com/QUEAHREN/EventGait.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes EventGait, a dual-stream end-to-end framework for gait recognition from event camera streams. The dynamic stream employs a Mixture of Spiking Experts (MoSE) with diverse neuron constants to model motion, while the static stream uses Cross-modal Structure Alignment (CroSA) to learn shape representations aligned with large vision models. To overcome the lack of event-based gait data, the authors introduce a synthesis pipeline converting RGB gait videos into event streams and release two new benchmarks (SUSTech1K-E and CCGR-Mini-E). They claim that event-based recognition matches camera-based performance under normal lighting but significantly outperforms it in low-light conditions, achieving new state-of-the-art results on both synthetic and real event-based gait benchmarks.

Significance. If the central claims hold after addressing validation gaps, the work would be a meaningful contribution to event-based vision for biometrics. It highlights practical advantages of event cameras (microsecond resolution, high dynamic range) for gait analysis in uncontrolled environments and provides open datasets and code that could support follow-on research. The empirical focus on low-light robustness addresses a known weakness of frame-based methods.

major comments (2)
  1. [Dataset construction and synthesis pipeline] The headline claim of significant outperformance in low-light scenarios rests on the synthesis pipeline (described in the dataset construction section) producing event streams whose temporal statistics, polarity distributions, and illumination-dependent noise match real event cameras. No quantitative validation (e.g., event-rate histograms or noise comparisons between synthesized SUSTech1K-E and real event recordings under matched low-light conditions) is provided; without this, the reported gains cannot be confidently attributed to event-camera properties rather than synthesis artifacts.
  2. [Experiments and results tables] Table reporting main results (likely Table 2 or equivalent in the experiments section): the cross-method comparison under low-light shows large margins, yet the manuscript does not report standard deviations across multiple runs or statistical significance tests. This makes it difficult to judge whether the claimed superiority is robust or sensitive to training details and post-hoc dataset choices.
minor comments (2)
  1. [Abstract and Experiments] The abstract states that the approach 'sets a new state of the art on both synthesized and real-world event-based gait benchmarks,' but the main text should explicitly list the prior event-based baselines and their scores for direct comparison.
  2. [Method, dynamic stream] Notation for the MoSE module (diverse neuron constants) is introduced without a clear equation or diagram showing how the mixture weights are computed; a small schematic would improve clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed comments, which help improve the rigor of our validation. We address each major comment below and will incorporate the suggested enhancements in the revised manuscript.

read point-by-point responses

  1. Referee: [Dataset construction and synthesis pipeline] The headline claim of significant outperformance in low-light scenarios rests on the synthesis pipeline (described in the dataset construction section) producing event streams whose temporal statistics, polarity distributions, and illumination-dependent noise match real event cameras. No quantitative validation (e.g., event-rate histograms or noise comparisons between synthesized SUSTech1K-E and real event recordings under matched low-light conditions) is provided; without this, the reported gains cannot be confidently attributed to event-camera properties rather than synthesis artifacts.

    Authors: We appreciate the referee highlighting the importance of quantitative validation for the synthesis pipeline. The pipeline follows established event generation models that incorporate illumination-dependent contrast thresholds and noise characteristics calibrated from real event cameras. To directly address this point, we will add event-rate histograms, polarity distribution comparisons, and noise analyses between the synthesized SUSTech1K-E data and real event recordings in the revised manuscript. We also note that our method demonstrates strong results on real-world event-based gait benchmarks, providing supporting evidence that the synthesized data preserves key event properties relevant to gait recognition. revision: yes

  2. Referee: [Experiments and results tables] Table reporting main results (likely Table 2 or equivalent in the experiments section): the cross-method comparison under low-light shows large margins, yet the manuscript does not report standard deviations across multiple runs or statistical significance tests. This makes it difficult to judge whether the claimed superiority is robust or sensitive to training details and post-hoc dataset choices.

    Authors: We agree that reporting variability and statistical significance strengthens the interpretation of the results. In the revised version, we will include standard deviations computed across multiple independent runs (with different random seeds) for the primary low-light comparison tables. We will additionally report results from statistical significance tests to quantify the robustness of the observed performance margins. These additions will clarify that the gains are not sensitive to specific training configurations. revision: yes

Circularity Check

0 steps flagged

Empirical contributions with independent experimental validation

full rationale

The paper is an empirical computer-vision study that introduces architectural components (MoSE dynamic stream, CroSA static stream) and a synthesis pipeline to create new benchmarks (SUSTech1K-E, CCGR-Mini-E). All performance claims rest on reported experimental results comparing event-based and camera-based gait recognition under normal and low-light conditions. No equations, predictions, or first-principles derivations are presented that reduce by construction to quantities fitted on the same data or to self-citations; the synthesis pipeline is a methodological choice whose fidelity is an external assumption rather than a definitional tautology. The work is therefore self-contained against external benchmarks and receives the default non-circularity finding.

Axiom & Free-Parameter Ledger

1 free parameters · 0 axioms · 0 invented entities

The central claim rests on the empirical superiority of the proposed dual-stream architecture and on the assumption that synthetic event data faithfully represents real event camera output. No explicit mathematical axioms or newly invented physical entities are introduced.

free parameters (1)
  • diverse neuron constants in MoSE
    Multiple time constants are chosen to handle complex motion and illumination; their specific values are design choices that affect dynamic stream performance.

pith-pipeline@v0.9.0 · 5805 in / 1276 out tokens · 34920 ms · 2026-05-22T06:53:08.340168+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

105 extracted references · 105 canonical work pages · 1 internal anchor

  1. [1]

    A 240 × 180 130 db 3µs latency global shutter spatiotemporal vision sensor.IEEE Journal of Solid-State Circuits, 49(10):2333–2341, 2014

    Christian Brandli, Raphael Berner, Minhao Yang, Shih-Chii Liu, and Tobi Delbruck. A 240 × 180 130 db 3µs latency global shutter spatiotemporal vision sensor.IEEE Journal of Solid-State Circuits, 49(10):2333–2341, 2014. 1, 2

    work page 2014

  2. [2]

    Event-guided person re-identification via sparse-dense complementary learning

    Chengzhi Cao, Xueyang Fu, Hongjian Liu, Yukun Huang, Kunyu Wang, Jiebo Luo, and Zheng-Jun Zha. Event-guided person re-identification via sparse-dense complementary learning. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17990– 17999, 2023. 2, 3, 5

    work page 2023

  3. [3]

    Learning robust event- guided representations for person re-identification: Cao et al.International Journal of Computer Vision, 134(2):82,

    Chengzhi Cao, Xueyang Fu, Senyan Xu, Chengjie Ge, Kunyu Wang, and Zheng-Jun Zha. Learning robust event- guided representations for person re-identification: Cao et al.International Journal of Computer Vision, 134(2):82,

    work page

  4. [4]

    Gaitset: Cross-view gait recognition through utilizing gait as a deep set.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(7):3467– 3478, 2022

    Hanqing Chao, Kun Wang, Yiwei He, Junping Zhang, and Jianfeng Feng. Gaitset: Cross-view gait recognition through utilizing gait as a deep set.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(7):3467– 3478, 2022. 6

    work page 2022

  5. [5]

    Efficient human pose estimation via 3d event point cloud

    Jiaan Chen, Hao Shi, Yaozu Ye, Kailun Yang, Lei Sun, and Kaiwei Wang. Efficient human pose estimation via 3d event point cloud. In2022 International Conference on 3D Vision (3DV), pages 1–10. IEEE, 2022. 3

    work page 2022

  6. [6]

    Segment any event streams via weighted adaptation of pivotal tokens

    Zhiwen Chen, Zhiyu Zhu, Yifan Zhang, Junhui Hou, Guangming Shi, and Jinjian Wu. Segment any event streams via weighted adaptation of pivotal tokens. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3890–3900, 2024. 3

    work page 2024

  7. [7]

    Qmambabsr: Burst image super-resolution with query state space model

    Xin Di, Long Peng, Peizhe Xia, Wenbo Li, Renjing Pei, Yang Cao, Yang Wang, and Zheng-Jun Zha. Qmambabsr: Burst image super-resolution with query state space model. InProceedings of the Computer Vision and Pattern Recog- nition Conference, pages 23080–23090, 2025. 3

    work page 2025

  8. [8]

    Metagait: Learning to learn an omni sample adaptive representation for gait recognition

    Huanzhang Dou, Pengyi Zhang, Wei Su, Yunlong Yu, and Xi Li. Metagait: Learning to learn an omni sample adaptive representation for gait recognition. InEuropean Conference on Computer Vision, pages 357–374. Springer, 2022. 2

    work page 2022

  9. [9]

    Gaitmpl: Gait recognition with memory-augmented progressive learning.IEEE Transac- tions on Image Processing, 2022

    Huanzhang Dou, Pengyi Zhang, Yuhan Zhao, Lin Dong, Zequn Qin, and Xi Li. Gaitmpl: Gait recognition with memory-augmented progressive learning.IEEE Transac- tions on Image Processing, 2022

    work page 2022

  10. [10]

    Gaitgci: Generative counterfactual intervention for gait recognition

    Huanzhang Dou, Pengyi Zhang, Wei Su, Yunlong Yu, Yin- ing Lin, and Xi Li. Gaitgci: Generative counterfactual intervention for gait recognition. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5578–5588, 2023. 2

    work page 2023

  11. [11]

    GaitPart: Temporal part-based model for gait recognition

    Chao Fan, Yunjie Peng, Chunshui Cao, Xu Liu, Saihui Hou, Jiannan Chi, Yongzhen Huang, Qing Li, and Zhiqiang He. GaitPart: Temporal part-based model for gait recognition. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14225–14233, 2020. 6

    work page 2020

  12. [12]

    Opengait: Revisiting gait recognition towards better practicality

    Chao Fan, Junhao Liang, Chuanfu Shen, Saihui Hou, Yongzhen Huang, and Shiqi Yu. Opengait: Revisiting gait recognition towards better practicality. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 9707–9716, 2023. 6

    work page 2023

  13. [13]

    Skeletongait: Gait recognition using skeleton maps

    Chao Fan, Jingzhe Ma, Dongyang Jin, Chuanfu Shen, and Shiqi Yu. Skeletongait: Gait recognition using skeleton maps. InProceedings of the AAAI conference on artificial intelligence, pages 1662–1669, 2024. 2, 6

    work page 2024

  14. [14]

    Opengait: A comprehensive benchmark study for gait recognition towards better practicality.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    Chao Fan, Saihui Hou, Junhao Liang, Chuanfu Shen, Jingzhe Ma, Dongyang Jin, Yongzhen Huang, and Shiqi Yu. Opengait: A comprehensive benchmark study for gait recognition towards better practicality.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025. 6

    work page 2025

  15. [15]

    Thomas Finateu, Atsumi Niwa, Daniel Matolin, Koya Tsuchimoto, Andrea Mascheroni, Etienne Reynaud, Poo- ria Mostafalu, Frederick Brady, Ludovic Chotard, Flo- rian LeGoff, Hirotsugu Takahashi, Hayato Wakabayashi, Yusuke Oike, and Christoph Posch. 5.10 a 1280×720 back-illuminated stacked temporal contrast event-based vi- sion sensor with 4.86µm pixels, 1.066...

    work page 2020

  16. [16]

    Event-driven heteroge- neous network for video deraining.International Journal of Computer Vision, 132(12):5841–5861, 2024

    Xueyang Fu, Chengzhi Cao, Senyan Xu, Fanrui Zhang, Kunyu Wang, and Zheng-Jun Zha. Event-driven heteroge- neous network for video deraining.International Journal of Computer Vision, 132(12):5841–5861, 2024. 1

    work page 2024

  17. [17]

    Gpgait: Generalized pose-based gait recognition.arXiv preprint arXiv:2303.05234, 2023

    Yang Fu, Shibei Meng, Saihui Hou, Xuecai Hu, and Yongzhen Huang. Gpgait: Generalized pose-based gait recognition.arXiv preprint arXiv:2303.05234, 2023. 2, 6

    work page arXiv 2023

  18. [18]

    Cut out the middle- man: Revisiting pose-based gait recognition

    Yang Fu, Saihui Hou, Shibei Meng, Xuecai Hu, Chunshui Cao, Xu Liu, and Yongzhen Huang. Cut out the middle- man: Revisiting pose-based gait recognition. InEuropean Conference on Computer Vision, pages 112–128. Springer,

    work page

  19. [19]

    Event- based, 6-dof camera tracking from photometric depth maps

    Guillermo Gallego, Jon EA Lund, Elias Mueggler, Henri Rebecq, Tobi Delbruck, and Davide Scaramuzza. Event- based, 6-dof camera tracking from photometric depth maps. IEEE transactions on pattern analysis and machine intelli- gence, 40(10):2402–2412, 2017. 1

    work page 2017

  20. [20]

    Eventmamba: En- hancing spatio-temporal locality with state space models for event-based video reconstruction

    Chengjie Ge, Xueyang Fu, Peng He, Kunyu Wang, Chengzhi Cao, and Zheng-Jun Zha. Eventmamba: En- hancing spatio-temporal locality with state space models for event-based video reconstruction. InProceedings of the AAAI Conference on Artificial Intelligence, pages 3104– 3112, 2025. 1

    work page 2025

  21. [21]

    Fractional-order spiking neural network

    Chengjie Ge, Yufeng Peng, Zihao Li, Qiyu Kang, Xueyang Fu, Xuhao Li, Qixin Zhang, Ren Junhao, and Zheng-Jun Zha. Fractional-order spiking neural network. InThe Four- teenth International Conference on Learning Representa- tions, 2026. 3

    work page 2026

  22. [22]

    Spiking neu- ral networks.International journal of neural systems, 19 (04):295–308, 2009

    Samanwoy Ghosh-Dastidar and Hojjat Adeli. Spiking neu- ral networks.International journal of neural systems, 19 (04):295–308, 2009. 3

    work page 2009

  23. [23]

    Camera-lidar cross-modality gait recognition

    Wenxuan Guo, Yingping Liang, Zhiyu Pan, Ziheng Xi, Jianjiang Feng, and Jie Zhou. Camera-lidar cross-modality gait recognition. InEuropean Conference on Computer Vi- sion, pages 439–455. Springer, 2024. 2

    work page 2024

  24. [24]

    Gait recognition in the wild: A large-scale bench- mark and nas-based baseline.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    Xianda Guo, Zheng Zhu, Tian Yang, Beibei Lin, Junjie Huang, Jiankang Deng, Guan Huang, Jie Zhou, and Jiwen Lu. Gait recognition in the wild: A large-scale bench- mark and nas-based baseline.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025. 2

    work page 2025

  25. [25]

    Gait recognition in large- scale free environment via single lidar

    Xiao Han, Yiming Ren, Peishan Cong, Yujing Sun, Jingya Wang, Lan Xu, and Yuexin Ma. Gait recognition in large- scale free environment via single lidar. InACM Multimedia 2024, 2024. 2, 6

    work page 2024

  26. [26]

    v2e: From video frames to realistic dvs events

    Yuhuang Hu, Shih-Chii Liu, and Tobi Delbruck. v2e: From video frames to realistic dvs events. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1312–1321, 2021. 5

    work page 2021

  27. [27]

    Occluded gait recognition with mixture of experts: an action detection perspective

    Panjian Huang, Yunjie Peng, Saihui Hou, Chunshui Cao, Xu Liu, Zhiqiang He, and Yongzhen Huang. Occluded gait recognition with mixture of experts: an action detection perspective. InEuropean Conference on Computer Vision, pages 380–397. Springer, 2024. 2

    work page 2024

  28. [28]

    V ocabulary-guided gait recognition

    Panjian Huang, Saihui Hou, Chunshui Cao, Xu Liu, and Yongzhen Huang. V ocabulary-guided gait recognition. In The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025. 6

    work page 2025

  29. [29]

    Learning a unified template for gait recognition

    Panjian Huang, Saihui Hou, Junzhou Huang, and Yongzhen Huang. Learning a unified template for gait recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 12459–12469, 2025. 2, 6

    work page 2025

  30. [30]

    Context-sensitive temporal feature learning for gait recog- nition

    Xiaohu Huang, Duowang Zhu, Hao Wang, Xinggang Wang, Bo Yang, Botao He, Wenyu Liu, and Bin Feng. Context-sensitive temporal feature learning for gait recog- nition. InProceedings of the IEEE/CVF international con- ference on computer vision, pages 12909–12918, 2021

    work page 2021

  31. [31]

    Condition-adaptive graph convolution learning for skeleton-based gait recogni- tion.IEEE Transactions on Image Processing, 32:4773– 4784, 2023

    Xiaohu Huang, Xinggang Wang, Zhidianqiu Jin, Bo Yang, Botao He, Bin Feng, and Wenyu Liu. Condition-adaptive graph convolution learning for skeleton-based gait recogni- tion.IEEE Transactions on Image Processing, 32:4773– 4784, 2023. 2, 6

    work page 2023

  32. [32]

    Rbs- former: Enhanced transformer network for raw image super-resolution

    Siyuan Jiang, Senyan Xu, and Xingfu Wang. Rbs- former: Enhanced transformer network for raw image super-resolution. InProceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, pages 6479–6488, 2024. 3

    work page 2024

  33. [33]

    Ex- ploring more from multiple gait modalities for human iden- tification.arXiv preprint arXiv:2412.11495, 2024

    Dongyang Jin, Chao Fan, Weihua Chen, and Shiqi Yu. Ex- ploring more from multiple gait modalities for human iden- tification.arXiv preprint arXiv:2412.11495, 2024. 6

    work page arXiv 2024

  34. [34]

    On denoising walking videos for gait recognition

    Dongyang Jin, Chao Fan, Jingzhe Ma, Jingkai Zhou, Wei- hua Chen, and Shiqi Yu. On denoising walking videos for gait recognition. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 12347–12357,

    work page

  35. [35]

    Event-guided unified framework for low-light video enhancement, frame inter- polation, and deblurring

    Taewoo Kim and Kuk-Jin Yoon. Event-guided unified framework for low-light video enhancement, frame inter- polation, and deblurring. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 8524– 8534, 2025. 3, 5

    work page 2025

  36. [36]

    Frequency- aware event-based video deblurring for real-world motion blur

    Taewoo Kim, Hoonhee Cho, and Kuk-Jin Yoon. Frequency- aware event-based video deblurring for real-world motion blur. InProceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, pages 24966–24976,

    work page

  37. [37]

    Be- yond sparse keypoints: Dense pose modeling for robust gait recognition

    Wenpeng Lang, Saihui Hou, and Yongzhen Huang. Be- yond sparse keypoints: Dense pose modeling for robust gait recognition. InProceedings of the 33rd ACM International Conference on Multimedia, pages 669–678, 2025. 2

    work page 2025

  38. [38]

    Fouriermamba: Fourier learning integration with state space models for image deraining.arXiv preprint arXiv:2405.19450, 2024

    Dong Li, Yidi Liu, Xueyang Fu, Senyan Xu, and Zheng- Jun Zha. Fouriermamba: Fourier learning integration with state space models for image deraining.arXiv preprint arXiv:2405.19450, 2024. 3

    work page arXiv 2024

  39. [39]

    Re- thinking appearance-based deep gait recognition: Reviews, analysis, and insights from gait recognition evolution.IEEE Transactions on Neural Networks and Learning Systems,

    Jingqi Li, Yuzhen Zhang, Yi Zeng, Changxin Ye, Wenzheng Xu, Xianye Ben, Fei-Yue Wang, and Junping Zhang. Re- thinking appearance-based deep gait recognition: Reviews, analysis, and insights from gait recognition evolution.IEEE Transactions on Neural Networks and Learning Systems,

    work page

  40. [40]

    3d feature tracking via event camera

    Siqi Li, Zhikuan Zhou, Zhou Xue, Yipeng Li, Shaoyi Du, and Yue Gao. 3d feature tracking via event camera. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18974–18983, 2024. 3

    work page 2024

  41. [41]

    End-to-end model-based gait recog- nition

    Xiang Li, Yasushi Makihara, Chi Xu, Yasushi Yagi, Shiqi Yu, and Mingwu Ren. End-to-end model-based gait recog- nition. InProceedings of the Asian Conference on Com- puter Vision, 2020. 2

    work page 2020

  42. [42]

    Multi-view large population gait database with human meshes and its performance evaluation.IEEE Transactions on Biometrics, Behavior, and Identity Science, 4(2):234– 248, 2022

    Xiang Li, Yasushi Makihara, Chi Xu, and Yasushi Yagi. Multi-view large population gait database with human meshes and its performance evaluation.IEEE Transactions on Biometrics, Behavior, and Identity Science, 4(2):234– 248, 2022. 2

    work page 2022

  43. [43]

    Depthgait: Multi-scale cross- level feature fusion of rgb-derived depth and silhouette se- quences for robust gait recognition

    Xinzhu Li, Juepeng Zheng, Yikun Chen, Xudong Mao, Guanghui Yue, Wei Zhou, Chenlei Lv, Ruomei Wang, Fan Zhou, and Baoquan Zhao. Depthgait: Multi-scale cross- level feature fusion of rgb-derived depth and silhouette se- quences for robust gait recognition. InProceedings of the 33rd ACM International Conference on Multimedia, pages 2333–2341, 2025. 6

    work page 2025

  44. [44]

    End-to-end model-based gait recognition using syn- chronized multi-view pose constraint

    Li, Xiang and Makihara, Yasushi and Xu, Chi and Yagi, Ya- sushi. End-to-end model-based gait recognition using syn- chronized multi-view pose constraint. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 4106–4115, 2021. 2

    work page 2021

  45. [45]

    Towards robust event-guided low- light image enhancement: A large-scale real-world event- image dataset and novel approach

    Guoqiang Liang, Kanghao Chen, Hangyu Li, Yunfan Lu, and Lin Wang. Towards robust event-guided low- light image enhancement: A large-scale real-world event- image dataset and novel approach. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 23–33, 2024. 3, 5

    work page 2024

  46. [46]

    Gaitedge: Beyond plain end-to-end gait recognition for better practicality

    Junhao Liang, Chao Fan, Saihui Hou, Chuanfu Shen, Yongzhen Huang, and Shiqi Yu. Gaitedge: Beyond plain end-to-end gait recognition for better practicality. InCom- puter Vision – ECCV 2022, 2022. 2

    work page 2022

  47. [47]

    Coherent event guided low-light video enhancement

    Jinxiu Liang, Yixin Yang, Boyu Li, Peiqi Duan, Yong Xu, and Boxin Shi. Coherent event guided low-light video enhancement. InProceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 10615– 10625, 2023. 3

    work page 2023

  48. [48]

    Pose-based temporal-spatial network (ptsn) for gait recognition with carrying and clothing vari- ations

    Rijun Liao, Chunshui Cao, Edel B Garcia, Shiqi Yu, and Yongzhen Huang. Pose-based temporal-spatial network (ptsn) for gait recognition with carrying and clothing vari- ations. InChinese conference on biometric recognition, pages 474–483. Springer, 2017. 2

    work page 2017

  49. [49]

    Gait recognition with multiple-temporal-scale 3d convolutional neural net- work

    Beibei Lin, Shunli Zhang, and Feng Bao. Gait recognition with multiple-temporal-scale 3d convolutional neural net- work. InProceedings of the 28th ACM international con- ference on multimedia, pages 3054–3062, 2020. 2

    work page 2020

  50. [50]

    Gait recognition via effective global-local feature representation and local temporal aggregation

    Beibei Lin, Shunli Zhang, and Xin Yu. Gait recognition via effective global-local feature representation and local temporal aggregation. InProceedings of the IEEE/CVF In- ternational Conference on Computer Vision, pages 14648– 14656, 2021. 2

    work page 2021

  51. [51]

    Seeing motion at nighttime with an event camera

    Haoyue Liu, Shihan Peng, Lin Zhu, Yi Chang, Hanyu Zhou, and Luxin Yan. Seeing motion at nighttime with an event camera. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 25648– 25658, 2024. 4

    work page 2024

  52. [52]

    Event-conditioned dual-modal fu- sion for motion deblurring

    Kean Liu, Mingchen Zhong, Senyan Xu, Zhijing Sun, Ji- aying Zhu, Chengjie Ge, Xingbo Wang, Xin Lu, Xueyang Fu, and Zheng-Jun Zha. Event-conditioned dual-modal fu- sion for motion deblurring. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 1482– 1492, 2025. 3

    work page 2025

  53. [53]

    Dreamuhd: Frequency enhanced variational autoencoder for ultra-high-definition image restoration

    Yidi Liu, Dong Li, Jie Xiao, Yuanfei Bao, Senyan Xu, and Xueyang Fu. Dreamuhd: Frequency enhanced variational autoencoder for ultra-high-definition image restoration. In Proceedings of the AAAI Conference on Artificial Intelli- gence, pages 5712–5720, 2025. 3

    work page 2025

  54. [54]

    Evenformer: Dynamic even transformer for real- world image restoration

    Xin Lu, Yuanfei Bao, Jiarong Yang, Anya Hu, Jie Xiao, Kunyu Wang, Dong Li, Senyan Xu, Kean Liu, Xueyang Fu, et al. Evenformer: Dynamic even transformer for real- world image restoration. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 1081– 1091, 2025. 3

    work page 2025

  55. [55]

    Mixture of ex- perts: a literature survey.Artificial Intelligence Review, 42 (2):275–293, 2014

    Saeed Masoudnia and Reza Ebrahimpour. Mixture of ex- perts: a literature survey.Artificial Intelligence Review, 42 (2):275–293, 2014. 4

    work page 2014

  56. [56]

    From fastposegait to gpgait++: Bridging the past and future for pose-based gait recognition.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    Shibei Meng, Yang Fu, Saihui Hou, Xuecai Hu, Chunshui Cao, Xu Liu, and Yongzhen Huang. From fastposegait to gpgait++: Bridging the past and future for pose-based gait recognition.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025. 2

    work page 2025

  57. [57]

    Seeing from magic mirror: Contrastive learning from reconstruction for pose-based gait recognition

    Shibei Meng, Saihui Hou, Yang Fu, Xuecai Hu, Junzhou Huang, and Yongzhen Huang. Seeing from magic mirror: Contrastive learning from reconstruction for pose-based gait recognition. InProceedings of the 33rd ACM Interna- tional Conference on Multimedia, pages 7719–7728, 2025. 2

    work page 2025

  58. [58]

    Data-driven feature tracking for event cameras

    Nico Messikommer, Carter Fang, Mathias Gehrig, and Da- vide Scaramuzza. Data-driven feature tracking for event cameras. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5642– 5651, 2023. 3

    work page 2023

  59. [59]

    Automatic recognition by gait.Proceedings of the IEEE, 94(11):2013–2024, 2006

    Mark S Nixon and John N Carter. Automatic recognition by gait.Proceedings of the IEEE, 94(11):2013–2024, 2006. 1

    work page 2013

  60. [60]

    DINOv2: Learning Robust Visual Features without Supervision

    Maxime Oquab, Timoth ´ee Darcet, Th ´eo Moutakanni, Huy V o, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, et al. Dinov2: Learning robust visual features without supervi- sion.arXiv preprint arXiv:2304.07193, 2023. 5

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  61. [61]

    Efficient real-world image super-resolution via adaptive directional gradient convolution.arXiv preprint arXiv:2405.07023, 2024

    Long Peng, Yang Cao, Renjing Pei, Wenbo Li, Ji- aming Guo, Xueyang Fu, Yang Wang, and Zheng-Jun Zha. Efficient real-world image super-resolution via adaptive directional gradient convolution.arXiv preprint arXiv:2405.07023, 2024. 3

    work page arXiv 2024

  62. [62]

    Lightweight adaptive feature de-drifting for compressed image classification.IEEE Transactions on Multimedia, 26: 6424–6436, 2024

    Long Peng, Yang Cao, Yuejin Sun, and Yang Wang. Lightweight adaptive feature de-drifting for compressed image classification.IEEE Transactions on Multimedia, 26: 6424–6436, 2024

    work page 2024

  63. [63]

    Unveiling hidden details: A raw data-enhanced paradigm for real-world super-resolution.arXiv preprint arXiv:2411.10798, 2024

    Long Peng, Wenbo Li, Jiaming Guo, Xin Di, Haoze Sun, Yong Li, Renjing Pei, Yang Wang, Yang Cao, and Zheng- Jun Zha. Unveiling hidden details: A raw data-enhanced paradigm for real-world super-resolution.arXiv preprint arXiv:2411.10798, 2024

    work page arXiv 2024

  64. [64]

    Towards real- istic data generation for real-world super-resolution.arXiv preprint arXiv:2406.07255, 2024

    Long Peng, Wenbo Li, Renjing Pei, Jingjing Ren, Jiaqi Xu, Yang Wang, Yang Cao, and Zheng-Jun Zha. Towards real- istic data generation for real-world super-resolution.arXiv preprint arXiv:2406.07255, 2024

    work page arXiv 2024

  65. [65]

    Directing mamba to complex textures: An effi- cient texture-aware state space model for image restoration

    Long Peng, Xin Di, Zhanfeng Feng, Wenbo Li, Renjing Pei, Yang Wang, Xueyang Fu, Yang Cao, and Zheng- Jun Zha. Directing mamba to complex textures: An effi- cient texture-aware state space model for image restoration. arXiv preprint arXiv:2501.16583, 2025

    work page arXiv 2025

  66. [66]

    Boosting image de-raining via central-surrounding synergistic convolution

    Long Peng, Yang Wang, Xin Di, Xueyang Fu, Yang Cao, Zheng-Jun Zha, et al. Boosting image de-raining via central-surrounding synergistic convolution. InProceed- ings of the AAAI Conference on Artificial Intelligence, pages 6470–6478, 2025

    work page 2025

  67. [67]

    Learning rich features for gait recognition by integrating skeletons and silhouettes.Multimedia Tools and Applica- tions, pages 1–22, 2023

    Yunjie Peng, Kang Ma, Yang Zhang, and Zhiqiang He. Learning rich features for gait recognition by integrating skeletons and silhouettes.Multimedia Tools and Applica- tions, pages 1–22, 2023

    work page 2023

  68. [68]

    Learning rich features for gait recognition by integrating skeletons and silhouettes.Multimedia Tools and Applica- tions, 83(3):7273–7294, 2024

    Yunjie Peng, Kang Ma, Yang Zhang, and Zhiqiang He. Learning rich features for gait recognition by integrating skeletons and silhouettes.Multimedia Tools and Applica- tions, 83(3):7273–7294, 2024. 3, 6

    work page 2024

  69. [69]

    Learn- ing transferable visual models from natural language super- vision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learn- ing transferable visual models from natural language super- vision. InInternational conference on machine learning, pages 8748–8763. PmLR, 2021. 5

    work page 2021

  70. [70]

    Oumvlp-of: Multi-view large population gait database with dense op- tical flow and its performance evaluation.IEEE Access,

    Allam Shehata, Francisco M Castro, Nicol ´as Guil, Manuel J Mar ´ın-Jim´enez, and Yasushi Yagi. Oumvlp-of: Multi-view large population gait database with dense op- tical flow and its performance evaluation.IEEE Access,

    work page

  71. [71]

    Lidargait: Benchmarking 3d gait recognition with point clouds

    Chuanfu Shen, Chao Fan, Wei Wu, Rui Wang, George Q Huang, and Shiqi Yu. Lidargait: Benchmarking 3d gait recognition with point clouds. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1054–1063, 2023. 1, 2, 5, 6, 7, 8

    work page 2023

  72. [72]

    Gait recognition with mask-based regularization

    Chuanfu Shen, Beibei Lin, Shunli Zhang, Xin Yu, George Q Huang, and Shiqi Yu. Gait recognition with mask-based regularization. In2023 IEEE International Joint Confer- ence on Biometrics (IJCB), pages 1–10. IEEE, 2023. 2

    work page 2023

  73. [73]

    A comprehensive survey on deep gait recognition: Algorithms, datasets, and challenges.IEEE Transactions on Biometrics, Behavior, and Identity Science,

    Chuanfu Shen, Shiqi Yu, Jilong Wang, George Q Huang, and Liang Wang. A comprehensive survey on deep gait recognition: Algorithms, datasets, and challenges.IEEE Transactions on Biometrics, Behavior, and Identity Science,

    work page

  74. [74]

    Li- dargait++: Learning local features and size awareness from lidar point clouds for 3d gait recognition

    Chuanfu Shen, Rui Wang, Lixin Duan, and Shiqi Yu. Li- dargait++: Learning local features and size awareness from lidar point clouds for 3d gait recognition. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 6627–6636, 2025. 1, 2, 6, 7, 8

    work page 2025

  75. [75]

    Yunjae Suh, Seungnam Choi, Masamichi Ito, Jeongseok Kim, Youngho Lee, Jongseok Seo, Heejae Jung, Dong- Hee Yeo, Seol Namgung, Jongwoo Bong, Sehoon Yoo, Seung-Hun Shin, Doowon Kwon, Pilkyu Kang, Seokho Kim, Hoonjoo Na, Kihyun Hwang, Changwoo Shin, Jun- Seok Kim, Paul K. J. Park, Joonseok Kim, Hyunsurk Ryu, and Yongin Park. A 1280×960 dynamic vision sensor ...

    work page 2020

  76. [76]

    Evdm: Event-based real-world video deblurring with mamba

    Zhijing Sun, Senyan Xu, Kean Liu, Runze Tian, Xueyang Fu, and Zheng-Jun Zha. Evdm: Event-based real-world video deblurring with mamba. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 13793–13803, 2025. 1

    work page 2025

  77. [77]

    Deep learning in spiking neural networks.Neural networks, 111:47–63, 2019

    Amirhossein Tavanaei, Masoud Ghodrati, Saeed Reza Kheradpisheh, Timoth´ee Masquelier, and Anthony Maida. Deep learning in spiking neural networks.Neural networks, 111:47–63, 2019. 4

    work page 2019

  78. [78]

    Towards a deeper under- standing of skeleton-based gait recognition

    Torben Teepe, Johannes Gilg, Fabian Herzog, Stefan H¨ormann, and Gerhard Rigoll. Towards a deeper under- standing of skeleton-based gait recognition. InProceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 1569–1577, 2022. 6

    work page 2022

  79. [79]

    Causal interven- tion for sparse-view gait recognition

    Jilong Wang, Saihui Hou, Yan Huang, Chunshui Cao, Xu Liu, Yongzhen Huang, and Liang Wang. Causal interven- tion for sparse-view gait recognition. InProceedings of the 31st ACM International Conference on Multimedia, pages 77–85, 2023. 2

    work page 2023

  80. [80]

    Free lunch for gait recognition: A novel relation descriptor

    Jilong Wang, Saihui Hou, Yan Huang, Chunshui Cao, Xu Liu, Yongzhen Huang, Tianzhu Zhang, and Liang Wang. Free lunch for gait recognition: A novel relation descriptor. InEuropean Conference on Computer Vision, pages 39–56. Springer, 2024

    work page 2024

Showing first 80 references.