Pith. sign in

REVIEW 3 major objections 2 minor 71 references

MPT: Motion Prompt Tuning for Micro-Expression Recognition

T0 review · 3 major / 2 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper claims that turning subtle facial motions into prompts lets large pre-trained models recognize micro-expressions, and that this beats state-of-the-art methods on three widely used datasets.

desk verdict A plausible prompt-tuning method for micro-expression recognition whose SOTA claim I cannot check because the supplied full text is corrupted; worth a proper look, not a desk reject. read the letter →

arxiv 2508.09446 v1 pith:A5QBMJRF submitted 2025-08-13 cs.CV

Jiateng Liu , Hengcan Shi , Feng Chen , Zhiwen Shao , Yaonan Wang , Jianfei Cai , Wenming Zheng This is my paper
classification cs.CV
keywords micro-expressionrecognitionmotionprompttuninglargepre-trainedmodelsmagnificationGaussiantokenizationgroupadapterparameter-efficientfine-tuningaffectivecomputing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Large pre-trained models have rich visual representations but miss micro-expressions because the discriminating motion is tiny and brief. The paper introduces Motion Prompt Tuning (MPT), which magnifies subtle facial motion, converts it into Gaussian-smoothed tokens, and feeds those motion prompts to a large pre-trained model together with a small group adapter. On three widely used micro-expression recognition datasets, the authors report that MPT consistently outperforms previous methods. If correct, this offers a parameter-efficient way to adapt large models to tasks where the signal is very small and annotated data are scarce.

What carries the argument

Motion prompt generation, a two-step process: motion magnification amplifies the small facial displacements that characterize micro-expressions, and Gaussian tokenization converts the magnified motion into a set of soft tokens that are fed into a large pre-trained vision transformer. A group adapter is inserted into the model to refine the target-domain representation, helping the model separate micro-expression classes that look similar. The machinery's job is to transform subtle, noisy motion into a representation that pre-trained layers can attend to and classify.

What would settle it

Run the MPT pipeline on one of the three datasets with the motion prompt generation disabled or with the magnification factor set to 1. If accuracy does not drop materially, the motion prompts are not carrying the claimed load. A second check: sweep the magnification factor over a wide range; if performance is flat, the specific amplification step is not doing the claimed work.

Watch

Extended reading notes

Core claim

The central claim is that subtle motion in micro-expressions can be turned into prompts that a large pre-trained model can use, rather than trying to learn the motion from scratch or simply fine-tuning the whole model. MPT first magnifies inter-frame motion to make micro-movements visible, then applies Gaussian tokenization to turn the magnified motion field into soft prompt tokens. A group adapter inserted into the transformer further adapts the model to the micro-expression domain. The paper reports that this design consistently surpasses state-of-the-art results on three widely used MER datasets, arguing that the main bottleneck is not model capacity but the format in which subtle motion

Load-bearing premise

The reported gains rest on the assumption that motion magnification plus Gaussian tokenization isolates the subtle, class-relevant motion cues while suppressing noise, and that the chosen magnification and tokenization settings transfer across subjects and datasets.

Editorial extensions

If this is right

  • If MPT works as reported, large pre-trained vision models can be adapted to micro-expression recognition with only light-weight prompts and a small adapter, reducing the need for large annotated micro-expression datasets.
  • The motion-prompt design provides a template for injecting domain-specific motion knowledge into frozen transformers, potentially extending to other subtle-motion recognition tasks.
  • Consistent gains across three datasets suggest that magnification-plus-Gaussian-tokenization is a generally useful preprocessing step for MER, not a trick tied to one benchmark.
  • Because only prompts and the adapter are learned, MPT can scale to larger pre-trained backbones without a proportional increase in task-specific parameters.
  • The approach could make micro-expression recognition more practical in low-annotation settings such as medical or security screening, where collecting expert-labeled micro-expressions is expensive.

Reading between the lines

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

  • A natural extension not explored in the paper: apply MPT's motion prompting to other subtle motion tasks like micro-gesture recognition or subtle physiological signal analysis, and test whether the same magnification and tokenization parameters transfer.
  • The magnification factor is likely dataset-dependent; if so, an automatic magnification selection or a multi-scale motion-prompt bank would be a natural next step, but the paper does not establish this.
  • Because prompt tuning is parameter-efficient, MPT could be combined with self-supervised pre-training on unlabeled facial videos to further close the annotation gap—an untested direction.
  • Gaussian tokenization treats motion as a soft distribution over locations rather than hard patches, which may be why the prompts remain robust to noise; this hypothesis is my inference, not the paper's claim.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 2 minor

Summary. The manuscript proposes Motion Prompt Tuning (MPT) for micro-expression recognition (MER). The abstract states that MPT adapts large pre-trained models by generating 'motion prompts' via motion magnification and Gaussian tokenization, and by inserting a group adapter to enhance target-domain representations. The claimed result is consistent state-of-the-art performance on three standard MER datasets. In the supplied source, however, only the abstract and a few garbled fragments are legible; the body is corrupted by an encoding failure. I therefore could not inspect the method description, the hyperparameters, the ablation results, the comparison tables, or the evaluation protocols.

Significance. If the empirical claim is true, MPT would be a useful parameter-efficient method for MER, where labels are scarce and subtle motion is essential. The high-level idea is coherent and addresses a real problem: adapting large pre-trained models to subtle-motion recognition. The paper attempts to validate on three widely used benchmarks and proposes a prompt-based mechanism that could be of interest to the MER community. However, the contribution is fundamentally empirical, and no credible experimental evidence is accessible in the submitted file. There is no readable ablation, no numerical comparison, and no code or machine-checked derivation. Thus the significance cannot be established from this submission.

major comments (3)
  1. [Abstract and full text] The central claim—'consistently surpasses state-of-the-art approaches'—is unverifiable because the supplied full text is mojibake. Every section after the abstract is corrupted, so no metric, confidence interval, baseline protocol, or comparison table can be read. This is not a presentation issue; the claim is load-bearing and cannot be checked as submitted.
  2. [Method description (unreadable)] The method relies on several hyperparameters and design choices: the motion magnification factor, the Gaussian tokenization scale, the group-adapter bottleneck dimension, and the prompt length. None of these are specified in the abstract, and the body text that should contain them is unreadable. Without these details the method cannot be reproduced, and the claimed superiority cannot be attributed to the proposed components rather than to particular hyperparameter choices.
  3. [Experimental protocol (unreadable)] The SOTA comparison requires knowing which three datasets were used, the exact evaluation protocol, the baseline versions, and whether hyperparameters were selected on the test sets. The corrupted tables and text do not reveal any of this. Even the table captions and numeric entries are unreadable, so I cannot determine whether the comparison is apples-to-apples with prior work.
minor comments (2)
  1. [Formatting] The document has severe encoding corruption throughout; for example, the arXiv identifier appears embedded mid-body, and equation symbols are replaced by replacement characters. A clean, readable version is needed before any substantive review.
  2. [Tables and equations] The ablations and comparison tables are not legible. If a readable version is provided, the authors should also ensure that table captions identify the dataset and metric, and that equations use standard notation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found; the abstract reports an empirical benchmark claim with no identifiable derivation that reduces to its own inputs.

full rationale

The supplied readable portion of the paper consists of the abstract; the full text is corrupted mojibake and cannot be parsed for equations, citations, or experimental tables. From the abstract, the derivation chain is: large pre-trained models fail to capture subtle facial movements; MPT introduces motion prompt generation (motion magnification and Gaussian tokenization) plus a group adapter; experiments on three widely used MER datasets show consistent improvement over state-of-the-art methods. None of these steps is self-definitional, and no fitted parameter is renamed as a prediction within the readable text. There are no visible self-citations or uniqueness theorems invoked to force the method. The central claim is an empirical comparison on external benchmarks, which is not circular by construction. The main unresolved risk is that magnification and tokenization hyperparameters may have been selected on the same test sets, but that cannot be checked from the available text and would be a correctness/overfitting concern rather than a circularity concern unless the full text revealed a direct reduction. Since no specific circular step can be quoted or exhibited, the appropriate finding is no significant circularity with score 0.

Assumptions & free parameters 4 free parameters · 3 assumptions · 2 invented entities

These parameters and components are the extra ingredients the paper adds. Without seeing the full text, the precise values and their sensitivity are unknown.

free parameters (4)
  • motion magnification factor
    Scales the subtle motion in the motion prompt generation stage; value not disclosed in abstract.
  • Gaussian tokenization kernel size/scale
    Controls the spatial smoothing/tokenization of motion prompts; value not disclosed.
  • group adapter bottleneck dimension
    Capacity of the inserted adapter; value not disclosed.
  • prompt length
    Number of learned prompt tokens for the large model; standard prompt tuning hyperparameter.
assumptions (3)
  • domain assumption Large pre-trained models provide general representations that can be effectively adapted to micro-expression recognition through prompts and adapters.
    The whole method is built on the belief that LMs are useful for MER if motion cues are supplied as prompts; stated in the introduction/abstract.
  • domain assumption Micro-expressions are characterized by subtle, transient facial movements that standard vision models do not capture, and motion magnification/tokenization can expose these movements.
    Motivates the motion prompt generation; if this false, the method's foundation fails.
  • domain assumption The three benchmark datasets and their evaluation protocols are comparable to those used in prior SOTA comparisons.
    The SOTA claim only holds under matched protocols; cannot be checked from abstract.
invented entities (2)
  • Motion prompts
    purpose: Input token sequences to the frozen LM that encode subtle facial motion information derived from magnification and Gaussian tokenization.
    They are validated only through the paper's experiments; no external falsifiable prediction beyond the reported datasets.
  • Group adapter
    purpose: Lightweight module inserted into the LM to adapt representations to the MER target domain.
    A new architectural component; no independent evidence outside the paper's evaluation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MPT: Motion Prompt Tuning for Micro-Expression Recognition." pith.science (2026). https://pith.science/paper/A5QBMJRF

@misc{pith2026250809446,
  author       = {Pith},
  title        = {Pith review of: MPT: Motion Prompt Tuning for Micro-Expression Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/A5QBMJRF}},
  note         = {Machine review of arXiv:2508.09446}
}
read the original abstract

Micro-expression recognition (MER) is crucial in the affective computing field due to its wide application in medical diagnosis, lie detection, and criminal investigation. Despite its significance, obtaining micro-expression (ME) annotations is challenging due to the expertise required from psychological professionals. Consequently, ME datasets often suffer from a scarcity of training samples, severely constraining the learning of MER models. While current large pre-training models (LMs) offer general and discriminative representations, their direct application to MER is hindered by an inability to capture transitory and subtle facial movements-essential elements for effective MER. This paper introduces Motion Prompt Tuning (MPT) as a novel approach to adapting LMs for MER, representing a pioneering method for subtle motion prompt tuning. Particularly, we introduce motion prompt generation, including motion magnification and Gaussian tokenization, to extract subtle motions as prompts for LMs. Additionally, a group adapter is carefully designed and inserted into the LM to enhance it in the target MER domain, facilitating a more nuanced distinction of ME representation. Furthermore, extensive experiments conducted on three widely used MER datasets demonstrate that our proposed MPT consistently surpasses state-of-the-art approaches and verifies its effectiveness.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

71 extracted references · 64 canonical work pages

  1. [1]

    Y. Bao, C. Wu, P. Zhang, C. Shan, Y. Qi, and X. Ben, ``Boosting micro-expression recognition via self-expression reconstruction and memory contrastive learning,'' IEEE Transactions on Affective Computing, 2024

    work page 2024

  2. [2]

    B. Song, K. Li, Y. Zong, J. Zhu, W. Zheng, J. Shi, and L. Zhao, ``Recognizing spontaneous micro-expression using a three-stream convolutional neural network,'' IEEE Access, vol. 7, pp. 184\,537--184\,551, 2019

    work page 2019

  3. [3]

    J. Hong, C. Lee, and H. Jung, ``Late fusion-based video transformer for facial micro-expression recognition,'' Applied Sciences, vol. 12, no. 3, p. 1169, 2022

    work page 2022

  4. [4]

    Zhang, X

    L. Zhang, X. Hong, O. Arandjelovi \'c , and G. Zhao, ``Short and long range relation based spatio-temporal transformer for micro-expression recognition,'' IEEE Transactions on Affective Computing, vol. 13, no. 4, pp. 1973--1985, 2022

    work page 1973

  5. [5]

    Krizhevsky, I

    A. Krizhevsky, I. Sutskever, and G. Hinton, ``Imagenet classification with deep convolutional neural networks,'' Advances in neural information processing systems, vol. 25, no. 2, 2012

    work page 2012

  6. [6]

    Simonyan and A

    K. Simonyan and A. Zisserman, ``Very deep convolutional networks for large-scale image recognition,'' Computer Science, 2014

    work page 2014

  7. [7]

    E. A. Haggard and K. S. Isaacs, ``Micromomentary facial expressions as indicators of ego mechanisms in psychotherapy,'' in Methods of research in psychotherapy. 1em plus 0.5em minus 0.4em Springer, 1966, pp. 154--165

    work page 1966

  8. [8]

    Y. Zong, X. Huang, W. Zheng, Z. Cui, and G. Zhao, ``Learning from hierarchical spatiotemporal descriptors for micro-expression recognition,'' IEEE Transactions on Multimedia, vol. 20, no. 11, pp. 3160--3172, 2018

    work page 2018

Show all 71 references
  1. [9]

    T. Zhou, J. Liu, Y. Jin, and L. Yao, ``Micromamba: State space model with partitioned window scan for micro-expression recognition,'' in Proceedings of the 6th ACM International Conference on Multimedia in Asia, 2024, pp. 1--7

    2024

  2. [10]

    J. Liu, Y. Zong, and W. Zheng, ``Cross-database micro-expression recognition based on transfer double sparse learning,'' Multimedia Tools and Applications, vol. 81, no. 30, pp. 43\,513--43\,530, 2022

    2022

  3. [11]

    Liong, R

    S. Liong, R. C. Phan, J. See, Y. Oh, and K. Wong, ``Optical strain based recognition of subtle emotions,'' in ISPACS, 2014. 1em plus 0.5em minus 0.4em IEEE , 2014, pp. 180--184

    2014

  4. [12]

    Y. Wang, J. See, W. Phan, and Y. H. Oh, ``Lbp with six intersection points: Reducing redundant information in lbp-top for micro-expression recognition,'' in Accv, 2015

    2015

  5. [13]

    A. C. L. Ngo, Y. Oh, R. C. Phan, and J. See, ``Eulerian emotion magnification for subtle expression recognition,'' in ICASSP. 1em plus 0.5em minus 0.4em IEEE , 2016, pp. 1243--1247

    2016

  6. [14]

    Zhao and M

    G. Zhao and M. Pietikainen, ``Dynamic texture recognition using local binary patterns with an application to facial expressions,'' IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 29, pp. p.915--928, 2007

    2007

  7. [15]

    M. Wei, X. Jiang, W. Zheng, Y. Zong, C. Lu, and J. Liu, ``Cmnet: contrastive magnification network for micro-expression recognition,'' in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 1, 2023, pp. 119--127

    2023

  8. [16]

    L. Lei, J. Li, T. Chen, and S. Li, ``A novel graph-tcn with a graph structured representation for micro-expression recognition,'' in Proceedings of the 28th ACM International Conference on Multimedia, 2020, pp. 2237--2245

    2020

  9. [17]

    M. Jia, L. Tang, B.-C. Chen, C. Cardie, S. Belongie, B. Hariharan, and S.-N. Lim, ``Visual prompt tuning,'' in Proceedings of the European Conference on Computer Vision. 1em plus 0.5em minus 0.4em Springer, 2022, pp. 709--727

    2022

  10. [18]

    Radford, J

    A. Radford, J. W. Kim et al., ``Learning transferable visual models from natural language supervision,'' in International conference on machine learning. 1em plus 0.5em minus 0.4em PMLR, 2021, pp. 8748--8763

    2021

  11. [19]

    H. Shi, M. Hayat, Y. Wu, and J. Cai, ``Proposalclip: Unsupervised open-category object proposal generation via exploiting clip cues,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 9611--9620

    2022

  12. [20]

    H. Shi, M. Hayat, and J. Cai, ``Transformer scale gate for semantic segmentation,'' in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 3051--3060

    2023

  13. [21]

    H. Shi, S. D. Dao, and J. Cai, ``Llmformer: Large language model for open-vocabulary semantic segmentation,'' International Journal of Computer Vision, pp. 1--18, 2024

    2024

  14. [22]

    M. U. Khattak, H. Rasheed, M. Maaz, S. Khan, and F. S. Khan, ``Maple: Multi-modal prompt learning,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 19\,113--19\,122

    2023

  15. [23]

    T. Yang, Y. Zhu, Y. Xie, A. Zhang, C. Chen, and M. Li, ``Aim: Adapting image models for efficient video action recognition,'' arXiv preprint arXiv:2302.03024, 2023

    2023 arXiv

  16. [24]

    J. Guo, H. Manukyan, C. Yang, C. Wang, L. Khachatryan, S. Navasardyan, S. Song, H. Shi, and G. Huang, ``Faceclip: Facial image-to-video translation via a brief text description,'' IEEE Transactions on Circuits and Systems for Video Technology, 2023

    2023

  17. [25]

    Houlsby, A

    N. Houlsby, A. Giurgiu, S. Jastrzebski, B. Morrone, Q. De Laroussilhe, A. Gesmundo, M. Attariyan, and S. Gelly, ``Parameter-efficient transfer learning for nlp,'' in International Conference on Machine Learning. 1em plus 0.5em minus 0.4em PMLR, 2019, pp. 2790--2799

    2019

  18. [26]

    P. Gao, S. Geng, R. Zhang, T. Ma, R. Fang, Y. Zhang, H. Li, and Y. Qiao, ``Clip-adapter: Better vision-language models with feature adapters,'' International Journal of Computer Vision, vol. 132, no. 2, pp. 581--595, 2024

    2024

  19. [27]

    Jiang, Y

    X. Jiang, Y. Zong, W. Zheng, J. Liu, and M. Wei, ``Seeking salient facial regions for cross-database micro-expression recognition,'' in 2022 26th International Conference on Pattern Recognition (ICPR). 1em plus 0.5em minus 0.4em IEEE, 2022, pp. 1019--1025

    2022

  20. [28]

    Liu, B.-J

    Y.-J. Liu, B.-J. Li, and Y.-K. Lai, ``Sparse mdmo: Learning a discriminative feature for spontaneous micro-expression recognition,'' IEEE Transactions on Affective Computing, 2018

    2018

  21. [29]

    S. Li, W. Zheng, Y. Zong, C. Lu, C. Tang, X. Jiang, J. Liu, and W. Xia, ``Bi-modality fusion for emotion recognition in the wild,'' in 2019 International Conference on Multimodal Interaction, 2019, pp. 589--594

    2019

  22. [30]

    M. Wei, W. Zheng, Y. Zong, X. Jiang, C. Lu, and J. Liu, ``A novel micro-expression recognition approach using attention-based magnification-adaptive networks,'' in ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). 1em plus 0.5e...

    2022

  23. [31]

    Verma, M

    M. Verma, M. Mandal, S. K. Reddy, Y. R. Meedimale, and S. K. Vipparthi, ``Efficient neural architecture search for emotion recognition,'' Expert Systems with Applications, vol. 224, p. 119957, 2023

    2023

  24. [32]

    Verma, S

    M. Verma, S. K. Vipparthi, G. Singh, and S. Murala, ``Learnet: Dynamic imaging network for micro expression recognition,'' IEEE Transactions on Image Processing, vol. 29, pp. 1618--1627, 2019

    2019

  25. [33]

    Z. Zhai, J. Zhao, C. Long, W. Xu, S. He, and H. Zhao, ``Feature representation learning with adaptive displacement generation and transformer fusion for micro-expression recognition,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, ...

    2023

  26. [34]

    X. Fan, X. Chen, M. Jiang, A. R. Shahid, and H. Yan, ``Selfme: Self-supervised motion learning for micro-expression recognition,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13\,834--13\,843

    2023

  27. [35]

    Y. Li, X. Huang, and G. Zhao, ``Joint local and global information learning with single apex frame detection for micro-expression recognition,'' IEEE Transactions on Image Processing, vol. 30, pp. 249--263, 2020

    2020

  28. [36]

    Chen, K.-H

    B. Chen, K.-H. Liu, Y. Xu, Q.-Q. Wu, and J.-F. Yao, ``Block division convolutional network with implicit deep features augmentation for micro-expression recognition,'' IEEE Transactions on Multimedia, vol. 25, pp. 1345--1358, 2022

    2022

  29. [37]

    Devlin, M.-W

    J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, ``Bert: Pre-training of deep bidirectional transformers for language understanding,'' arXiv preprint arXiv:1810.04805, 2018

    2018 arXiv

  30. [38]

    Dosovitskiy, L

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly et al., ``An image is worth 16x16 words: Transformers for image recognition at scale,'' arXiv preprint arXiv:2010.11929, 2020

    2010 arXiv

  31. [39]

    Carion, F

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, ``End-to-end object detection with transformers,'' in European conference on computer vision. 1em plus 0.5em minus 0.4em Springer, 2020, pp. 213--229

    2020

  32. [40]

    Y. Li, S. Xie, X. Chen, P. Dollar, K. He, and R. Girshick, ``Benchmarking detection transfer learning with vision transformers,'' arXiv preprint arXiv:2111.11429, 2021

    2021 arXiv

  33. [41]

    Y. Xing, Q. Wu, D. Cheng, S. Zhang, G. Liang, P. Wang, and Y. Zhang, ``Dual modality prompt tuning for vision-language pre-trained model,'' IEEE Transactions on Multimedia, 2023

    2023

  34. [42]

    J. Zhu, S. Lai, X. Chen, D. Wang, and H. Lu, ``Visual prompt multi-modal tracking,'' in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 9516--9526

    2023

  35. [43]

    G. Sun, C. Wang, Z. Zhang, J. Deng, S. Zafeiriou, and Y. Hua, ``Spatio-temporal prompting network for robust video feature extraction,'' in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 13\,587--13\,597

    2023

  36. [44]

    Z. Chen, Y. Duan, W. Wang, J. He, T. Lu, J. Dai, and Y. Qiao, ``Vision transformer adapter for dense predictions,'' arXiv preprint arXiv:2205.08534, 2022

    2022 arXiv

  37. [45]

    Y.-L. Sung, J. Cho, and M. Bansal, ``Vl-adapter: Parameter-efficient transfer learning for vision-and-language tasks,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 5227--5237

    2022

  38. [46]

    J. Pan, Z. Lin, X. Zhu, J. Shao, and H. Li, ``St-adapter: Parameter-efficient image-to-video transfer learning,'' NeurIPS, vol. 35, pp. 26\,462--26\,477, 2022

    2022

  39. [47]

    Bilen, B

    H. Bilen, B. Fernando, E. Gavves, and A. Vedaldi, ``Action recognition with dynamic image networks,'' IEEE transactions on pattern analysis and machine intelligence, vol. 40, no. 12, pp. 2799--2813, 2017

    2017

  40. [48]

    Y. Li, X. Huang, and G. Zhao, ``Can micro-expression be recognized based on single apex frame?'' in 2018 25th IEEE International Conference on Image Processing (ICIP). 1em plus 0.5em minus 0.4em IEEE, 2018, pp. 3094--3098

    2018

  41. [49]

    H.-Y. Wu, M. Rubinstein, E. Shih, J. Guttag, F. Durand, and W. Freeman, ``Eulerian video magnification for revealing subtle changes in the world,'' ACM transactions on graphics (TOG), vol. 31, no. 4, pp. 1--8, 2012

    2012

  42. [50]

    Y. S. Gan, S. Liong, W. Yau, Y. Huang, and T. L. Ken, ``Off-apexnet on micro-expression recognition system,'' Signal Process. Image Commun., vol. 74, pp. 129--139, 2019

    2019

  43. [51]

    Hendrycks and K

    D. Hendrycks and K. Gimpel, ``Gaussian error linear units (gelus),'' arXiv preprint arXiv:1606.08415, 2016

    2016 arXiv

  44. [52]

    D. H. Kim, W. J. Baddar, and Y. M. Ro, ``Micro-expression recognition with expression-state constrained spatio-temporal feature representations,'' in Acm on Multimedia Conference, 2016

    2016

  45. [53]

    H. Khor, J. See, S. Liong, R. C. Phan, and W. Lin, ``Dual-stream shallow networks for facial micro-expression recognition,'' in ICIP. 1em plus 0.5em minus 0.4em IEEE , 2019, pp. 36--40

    2019

  46. [54]

    B. Sun, S. Cao, D. Li, J. He, and L. Yu, ``Dynamic micro-expression recognition using knowledge distillation,'' IEEE Transactions on Affective Computing, 2020

    2020

  47. [55]

    Nguyen, C

    X.-B. Nguyen, C. N. Duong, X. Li, S. Gauch, H.-S. Seo, and K. Luu, ``Micron-bert: Bert-based facial micro-expression recognition,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 1482--1492

    2023

  48. [56]

    F. Wang, Y. Zong, J. Zhu, M. Wei, X. Xu, C. Lu, and W. Zheng, ``Progressively learning from macro-expressions for micro-expression recognition,'' in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). 1em plus 0.5em minus 0.4em I...

    2024

  49. [57]

    C. Li, R. Ba, X. Wang, M. Yu, X. Li, and D. Huang, ``Structure representation with adaptive and compact facial graph for micro-expression recognition,'' IEEE Transactions on Biometrics, Behavior, and Identity Science, 2024

    2024

  50. [58]

    J. Zhu, Y. Zong, J. Shi, C. Lu, H. Chang, and W. Zheng, ``Learning to rank onset-occurring-offset representations for micro-expression recognition,'' IEEE Transactions on Affective Computing, 2025

    2025

  51. [59]

    L. Lei, T. Chen, S. Li, and J. Li, ``Micro-expression recognition based on facial graph representation learning and facial action unit fusion,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 1571--1580

    2021

  52. [60]

    Huang and G

    X. Huang and G. Zhao, ``Spontaneous facial micro-expression analysis using spatiotemporal local radon-based binary pattern,'' in FADS, 2017

    2017

  53. [61]

    W.-J. Yan, X. Li, S.-J. Wang, G. Zhao, Y.-J. Liu, Y.-H. Chen, and X. Fu, ``Casme ii: An improved spontaneous micro-expression database and the baseline evaluation,'' PloS one, vol. 9, no. 1, 2014

    2014

  54. [62]

    A. K. Davison, C. Lansley, N. Costen, K. Tan, and M. H. Yap, ``Samm: A spontaneous micro-facial movement dataset,'' IEEE Transactions on Affective Computing, vol. 9, no. 1, pp. 116--129, 2016

    2016

  55. [63]

    X. Li, T. Pfister, X. Huang, G. Zhao, and M. Pietik \"a inen, ``A spontaneous micro-expression database: Inducement, collection and baseline,'' in 2013 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG). 1em plus 0.5em minus 0.4em I...

    2013

  56. [64]

    Zhang, ``Improved adam optimizer for deep neural networks,'' in 2018 IEEE/ACM 26th international symposium on quality of service (IWQoS)

    Z. Zhang, ``Improved adam optimizer for deep neural networks,'' in 2018 IEEE/ACM 26th international symposium on quality of service (IWQoS). 1em plus 0.5em minus 0.4em Ieee, 2018, pp. 1--2

    2018

  57. [65]

    B. Xia, W. Wang, S. Wang, and E. Chen, ``Learning from macro-expression: A micro-expression recognition framework,'' in Proceedings of the 28th ACM International Conference on Multimedia, 2020, pp. 2936--2944

    2020

  58. [66]

    Xia and S

    B. Xia and S. Wang, ``Micro-expression recognition enhanced by macro-expression from spatial-temporal domain.'' in IJCAI, 2021, pp. 1186--1193

    2021

  59. [67]

    C. Lee, J. Hong, and H. Jung, ``N-step pre-training and d \'e calcomanie data augmentation for micro-expression recognition,'' Sensors, vol. 22, no. 17, p. 6671, 2022

    2022

  60. [68]

    L. Zhou, Q. Mao, X. Huang, F. Zhang, and Z. Zhang, ``Feature refinement: An expression-specific feature learning and fusion method for micro-expression recognition,'' Pattern Recognition, vol. 122, p. 108275, 2022

    2022

  61. [69]

    Russakovsky, J

    O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy, A. Khosla, M. Bernstein et al., ``Imagenet large scale visual recognition challenge,'' International journal of computer vision, vol. 115, pp. 211--252, 2015

    2015

  62. [70]

    R. R. e. a. Selvaraju, ``Grad-cam: Visual explanations from deep networks via gradient-based localization,'' in Proceedings of the IEEE/CVF international conference on computer vision, 2017, pp. 618--626

    2017

  63. [71]

    Van der Maaten and G

    L. Van der Maaten and G. Hinton, ``Visualizing data using t-sne.'' Journal of machine learning research, vol. 9, no. 11, 2008

    2008

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.