pith. sign in

arxiv: 2606.11645 · v1 · pith:QDYWI5PTnew · submitted 2026-06-10 · 💻 cs.CV

Motion Reinforces Appearance: RGB-Skeleton Gated Residual Fusion for Micro-Gesture Online Recognition

Jialin Liu , Xinwen He , Pengyu Liu , Jiale Shi , Huaijuan Zang , Yanbin Hao This is my paper

Pith reviewed 2026-06-27 10:53 UTC · model grok-4.3

classification 💻 cs.CV
keywords micro-gesture recognitionRGB-skeleton fusiongated residual fusiononline action detectionmulti-modal temporal embeddingsDynamic TAD headSMG dataset
0
0 comments X

The pith

A gated residual module adaptively injects skeleton motion cues into RGB features to improve micro-gesture localization and classification in untrimmed videos.

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

The paper establishes that a dual-stream RGB-skeleton model can better capture spontaneous micro-gestures by projecting both modalities into shared multi-scale temporal embeddings and using a gated residual fusion to selectively add motion information to appearance features. This matters because single-modality approaches struggle with the subtle and variable nature of micro-gestures, which require precise start and end times plus category labels in continuous video. The fused representation is then passed to a Dynamic TAD head for online detection, yielding an F1 score of 40.88 on the SMG dataset and second place in the challenge track. The core mechanism avoids naive concatenation in favor of adaptive injection controlled by the gate.

Core claim

DyFADet+ extends the original DyFADet into a dual-stream RGB-skeleton framework in which both modalities are projected into shared multi-scale temporal embeddings and fused through a gated residual module that adaptively injects skeleton motion into the RGB representation; the fused features are decoded by a Dynamic TAD head for online classification and boundary regression, producing an F1 score of 40.88 on the SMG dataset.

What carries the argument

The gated residual module that adaptively injects skeleton motion into the RGB representation rather than using naive concatenation.

If this is right

  • The dual-stream model with shared temporal embeddings produces more accurate start-time, end-time, and category outputs for each micro-gesture instance than single-modality baselines.
  • Adaptive rather than fixed fusion allows the network to emphasize motion information only where it reinforces appearance cues.
  • The resulting features support real-time online processing on untrimmed video without requiring trimmed clips.
  • The same fused representation feeds directly into an existing Dynamic TAD decoding head for boundary regression.

Where Pith is reading between the lines

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

  • The gated injection pattern could be tested on other paired modalities such as depth or optical flow for the same task.
  • If the gate learns to suppress skeleton input on certain gesture classes, that pattern might reveal which micro-gestures are primarily appearance-driven.
  • Extending the shared embedding space to include audio could address cases where visual motion alone remains ambiguous.

Load-bearing premise

Skeleton motion supplies complementary cues that the gated residual module can usefully inject into RGB features, rather than performance gains arising mainly from architecture scale, training schedule, or dataset properties.

What would settle it

A controlled ablation in which the gated residual module is replaced by simple feature concatenation or by an RGB-only stream while keeping all other components identical and observing whether the F1 score remains at or above 40.88.

Figures

Figures reproduced from arXiv: 2606.11645 by Huaijuan Zang, Jiale Shi, Jialin Liu, Pengyu Liu, Xinwen He, Yanbin Hao.

Figure 1
Figure 1. Figure 1: Overview of the proposed DyFADet+ architecture for SMG micro-gesture detection. 3.3 RGB and Skeleton Feature The RGB branch adopts the projection architecture of DyFADet [43], employing a shallow temporal convolutional embedder followed by a sequence of Dynamic Encoder (DynE) blocks. By leveraging dynamic feature aggregation and progres￾sive temporal pooling, these blocks systematically downsample the sequ… view at source ↗
Figure 2
Figure 2. Figure 2: Gated residual fusion at a single pyramid level: skeleton features pass through an adapter, a sigmoid gate (conditioned on RGB and skeleton) selects motion cues, and the result is added to RGB with scaling α. modalities: G(ℓ) = σ  ϕℓ hF (ℓ) r , F (ℓ) s i , (1) F (ℓ) f = F (ℓ) r + α · G(ℓ) ⊙ ψℓ  F (ℓ) s  , (2) where the bracketed term [F (ℓ) r , F (ℓ) s ] concatenates the RGB and skeleton features alo… view at source ↗
read the original abstract

Micro-gesture analysis attracts increasing attention for inferring spontaneous emotion from subtle body movements. Micro-gesture online recognition, which localizes and classifies each gesture instance in untrimmed videos, is a core task in the 4th EI-MiGA-IJCAI Challenge. Compared with typical temporal action detection, MGR emphasizes the localization and classification of actions, requiring the model to output the start time, end time, and category of each micro-gesture. Moreover, since micro-gestures are highly spontaneous, relying solely on a single modality makes it difficult to capture the complete and accurate multi-modal cues. In this work, we propose DyFADet+, which extends DyFADet into a dual-stream RGB-skeleton framework. In our model, both modalities are projected into shared multi-scale temporal embeddings and fused through a gated residual module, which adaptively injects skeleton motion into the RGB representation rather than using naive concatenation. Finally, these fused features are decoded by a Dynamic TAD head for online classification and boundary regression. On the SMG dataset, our method achieves an F1 score of 40.88, ranking 2nd in the Micro-gesture Online Recognition track.

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 / 0 minor

Summary. The manuscript introduces DyFADet+, a dual-stream RGB-skeleton extension of DyFADet for micro-gesture online recognition. Both modalities are projected into shared multi-scale temporal embeddings and fused by a gated residual module that adaptively injects skeleton motion cues into the RGB stream (rather than naive concatenation); the fused representation is decoded by a Dynamic TAD head to produce online start/end times and class labels. On the SMG dataset the method reports an F1 score of 40.88 and second place in the Micro-gesture Online Recognition track of the 4th EI-MiGA-IJCAI Challenge.

Significance. If the gated residual fusion can be shown to be the operative factor, the work would supply a concrete, learnable mechanism for injecting complementary motion information into appearance features for subtle, spontaneous actions. The competitive ranking on an external challenge dataset indicates potential practical value for multi-modal micro-gesture analysis, but the absence of controlled ablations prevents assessment of whether the reported gain exceeds what would be obtained by simply increasing model capacity or altering the training schedule.

major comments (2)
  1. [Abstract] Abstract (paragraph on DyFADet+): the central claim that the gated residual module 'adaptively injects skeleton motion into the RGB representation rather than using naive concatenation' is load-bearing for the attribution of the 40.88 F1 score, yet the manuscript supplies no quantitative delta, ablation table, or controlled experiment that isolates the gated module while holding the dual-stream architecture, shared embeddings, and Dynamic TAD head fixed.
  2. [Abstract] Abstract: the reported F1 score of 40.88 is presented without baselines, ablation tables, error bars, or any description of how the gated module was trained or validated, so the performance claim cannot be checked against the DyFADet baseline or against alternative fusion strategies.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. The comments correctly identify that the manuscript does not currently provide ablations or baselines to isolate the contribution of the gated residual fusion. We address each point below and commit to adding the requested experiments in revision.

read point-by-point responses

  1. Referee: [Abstract] Abstract (paragraph on DyFADet+): the central claim that the gated residual module 'adaptively injects skeleton motion into the RGB representation rather than using naive concatenation' is load-bearing for the attribution of the 40.88 F1 score, yet the manuscript supplies no quantitative delta, ablation table, or controlled experiment that isolates the gated module while holding the dual-stream architecture, shared embeddings, and Dynamic TAD head fixed.

    Authors: We agree that the manuscript lacks a controlled ablation isolating the gated residual module. In the revised version we will add experiments that hold the dual-stream architecture, shared multi-scale embeddings, and Dynamic TAD head fixed while comparing gated residual fusion against naive concatenation and alternative fusion strategies, thereby supplying the requested quantitative deltas. revision: yes

  2. Referee: [Abstract] Abstract: the reported F1 score of 40.88 is presented without baselines, ablation tables, error bars, or any description of how the gated module was trained or validated, so the performance claim cannot be checked against the DyFADet baseline or against alternative fusion strategies.

    Authors: We acknowledge that the current manuscript does not include explicit baselines, ablation tables, error bars, or expanded training details for the gated module. The revision will incorporate comparisons to the original DyFADet, alternative fusion approaches, and a description of the training/validation protocol; error bars from repeated runs will be added if compute permits. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical result on external benchmark with independent evaluation

full rationale

The paper proposes DyFADet+ as an architectural extension (dual-stream RGB-skeleton with shared embeddings and gated residual fusion) and reports an F1 score of 40.88 on the SMG dataset from the EI-MiGA-IJCAI Challenge. No equations, predictions, or derivations are present that reduce by construction to fitted inputs or self-referential quantities. The central claim is an externally measured performance number on a held-out challenge dataset, which is falsifiable and does not rely on self-citation chains or ansatzes for its validity. Minor self-citation of the base DyFADet model (if by overlapping authors) is not load-bearing for the reported result.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no equations, derivations, or modeling assumptions are supplied that would allow enumeration of free parameters, axioms, or invented entities.

pith-pipeline@v0.9.1-grok · 5763 in / 1114 out tokens · 24075 ms · 2026-06-27T10:53:38.720913+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

45 extracted references · 4 canonical work pages

  1. [1]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Caba Heilbron, F., Escorcia, V., Ghanem, B., Carlos Niebles, J.: ActivityNet: A large-scale video benchmark for human activity understanding. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 961–970 (2015)

    2015

  2. [2]

    IEEE Transactions on Affective Computing1(1), 18–37 (2010).https://doi.org/10.1109/T-AFFC.2010.1 RGB-Skeleton Fusion for Micro-Gesture Recognition 11

    Calvo, R.A., D’Mello, S.: Affect detection: An interdisciplinary review of models, methods, and their applications. IEEE Transactions on Affective Computing1(1), 18–37 (2010).https://doi.org/10.1109/T-AFFC.2010.1 RGB-Skeleton Fusion for Micro-Gesture Recognition 11

    work page doi:10.1109/t-affc.2010.1 2010

  3. [3]

    arXiv preprint arXiv:2408.03097 (2024)

    Chen, G., Wang, F., Li, K., Wu, Z., Fan, H., Yang, Y., Wang, M., Guo, D.: Prototype learning for micro-gesture classification. arXiv preprint arXiv:2408.03097 (2024)

    arXiv 2024

  4. [4]

    In: Proceedings of the IJCAI 2024 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2024)

    Chen, H., Schuller, B.W., Adeli, E., Zhao, G.: The 2nd challenge on micro-gesture analysis for hidden emotion understanding (MiGA) 2024: Dataset and results. In: Proceedings of the IJCAI 2024 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2024). CEUR Workshop Proceedings, vol. 3848 (2024)

    2024

  5. [5]

    International Journal of Computer Vision131(5), 1346–1366 (2023).https://doi.org/10.1007/ s11263-023-01761-6

    Chen, H., Shi, H., Liu, X., Li, X., Zhao, G.: SMG: A micro-gesture dataset to- wards spontaneous body gestures for emotional stress state analysis. International Journal of Computer Vision131(5), 1346–1366 (2023).https://doi.org/10.1007/ s11263-023-01761-6

    2023

  6. [6]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Chen, Y., Zhang, Z., Yuan, C., Li, B., Deng, Y., Hu, W.: Channel-wise topology refinement graph convolution for skeleton-based action recognition. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 13359–13368 (2021)

    2021

  7. [7]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Duan, H., Zhao, Y., Chen, K., Lin, D., Dai, B.: Revisiting skeleton-based action recognition. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2969–2978 (2022)

    2022

  8. [8]

    PsycEXTRA Dataset (1969).https://doi.org/10.1037/e525532009-012

    Ekman, P., Friesen, W.V.: Nonverbal leakage and clues to deception. PsycEXTRA Dataset (1969).https://doi.org/10.1037/e525532009-012

    work page doi:10.1037/e525532009-012 1969

  9. [9]

    In: Proceedings of the 33rd ACM International Conference on Multimedia

    Gu, J., Li, K., Wang, F., Wei, Y., Wu, Z., Fan, H., Wang, M.: Motion matters: Motion-guided modulation network for skeleton-based micro-action recognition. In: Proceedings of the 33rd ACM International Conference on Multimedia. pp. 5461–5470 (2025)

    2025

  10. [10]

    arXiv preprint arXiv:2507.08344 (2025)

    Gu, J., Wang, F., Li, K., Wei, Y., Wu, Z., Guo, D.: MM-Gesture: Towards precise micro-gesture recognition through multimodal fusion. arXiv preprint arXiv:2507.08344 (2025)

    arXiv 2025

  11. [11]

    IEEE Transactions on Circuits and Systems for Video Technology (2024), arXiv:2403.05234

    Guo, D., Li, K., Hu, B., Zhang, Y., Wang, M.: Benchmarking micro-action recog- nition: Dataset, methods, and applications. IEEE Transactions on Circuits and Systems for Video Technology (2024), arXiv:2403.05234

    arXiv 2024

  12. [12]

    In: Proceed- ings of the IJCAI 2023 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2023)

    Guo, X.P., Peng, W., Huang, H., Xia, Z.: Micro-gesture online recognition with graph-convolution and multiscale transformers for long sequence. In: Proceed- ings of the IJCAI 2023 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2023). CEUR Workshop Proceedings, vol. 3522 (2023)

    2023

  13. [13]

    IEEE Transactions on Circuits and Systems for Video Technology32(10), 7120–7132 (2022)

    Hao, Y., Wang, S., Cao, P., Gao, X., Xu, T., Wu, J., He, X.: Attention in attention: Modeling context correlation for efficient video classification. IEEE Transactions on Circuits and Systems for Video Technology32(10), 7120–7132 (2022)

    2022

  14. [14]

    In: Proceedings of the ieee/cvf conference on computer vision and pattern recognition

    Hao, Y., Zhang, H., Ngo, C.W., He, X.: Group contextualization for video recog- nition. In: Proceedings of the ieee/cvf conference on computer vision and pattern recognition. pp. 928–938 (2022)

    2022

  15. [15]

    In: ECCV Workshop on Action Recognition with a Large Number of Classes (2014)

    Idrees, H., Zamir, A.R., Jiang, Y.G., Gorban, A., Laptev, I., Sukthankar, R., Shah, M.: THUMOS challenge: Action recognition with a large number of classes. In: ECCV Workshop on Action Recognition with a Large Number of Classes (2014)

    2014

  16. [16]

    arXiv preprint arXiv:2603.26586 (2026)

    Li, K., Gu, J., Wang, F., Wu, Z., Fan, H., Guo, D.: MA-Bench: Towards fine-grained micro-action understanding. arXiv preprint arXiv:2603.26586 (2026)

    arXiv 2026

  17. [17]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Li, K., Guo, D., Chen, G., Fan, C., Xu, J., Wu, Z., Fan, H., Wang, M.: Prototypical calibrating ambiguous samples for micro-action recognition. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 39, pp. 4815–4823 (2025)

    2025

  18. [18]

    In: Proceedings of the 31st ACM International Conference on Multimedia

    Li, K., Guo, D., Chen, G., Liu, F., Wang, M.: Data augmentation for human behavior analysis in multi-person conversations. In: Proceedings of the 31st ACM International Conference on Multimedia. pp. 9516–9520 (2023) 12 J. Liuet al

    2023

  19. [19]

    arXiv preprint arXiv:2307.10624 (2023)

    Li, K., Guo, D., Chen, G., Peng, X., Wang, M.: Joint skeletal and semantic embedding loss for micro-gesture classification. arXiv preprint arXiv:2307.10624 (2023)

    arXiv 2023

  20. [20]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (2025)

    Li, K., Liu, P., Guo, D., Wang, F., Wu, Z., Fan, H., Wang, M.: MMAD: Multi-label micro-action detection in videos. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (2025)

    2025

  21. [21]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Lin, C., Li, J., Wang, Y., Tai, Y., Luo, D., Cui, Z., Wang, C., Li, J., Huang, F., Ji, R.: Learning salient boundary feature for anchor-free temporal action localization. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 3320–3329 (2021)

    2021

  22. [22]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Lin, T., Liu, X., Li, X., Ding, E., Wen, S.: Bmn: Boundary-matching network for tem- poral action proposal generation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 3889–3898 (2019)

    2019

  23. [23]

    In: Proceedings of the IEEE International Conference on Computer Vision

    Lin, T.Y., Goyal, P., Girshick, R., He, K., Dollár, P.: Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision. pp. 2980–2988 (2017)

    2017

  24. [24]

    arXiv preprint arXiv:2503.15978 (2025)

    Liu, P., Dong, G., Guo, D., Li, K., Li, F., Yang, X., Wang, M., Ying, X.: A survey on fmri-based brain decoding for reconstructing multimodal stimuli. arXiv preprint arXiv:2503.15978 (2025)

    arXiv 2025

  25. [25]

    arXiv preprint arXiv:2507.09512 (2025)

    Liu, P., Li, K., Wang, F., Wei, Y., She, J., Guo, D.: Online micro-gesture recog- nition using data augmentation and spatial-temporal attention. arXiv preprint arXiv:2507.09512 (2025)

    arXiv 2025

  26. [26]

    In: Proceedings of the IJCAI 2024 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2024)

    Liu, P., Wang, F., Li, K., Chen, G., Wei, Y., Tang, S., Wu, Z., Guo, D.: Micro-gesture online recognition using learnable query points. In: Proceedings of the IJCAI 2024 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2024). CEUR Workshop Proceedings, vol. 3848 (2024)

    2024

  27. [27]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops

    Liu, S., Zhao, C., Zohra, F., Soldan, M., Pardo, A., Xu, M., Alssum, L., Ra- mazanova, M., Alcázar, J.L., Cioppa, A., Giancola, S., Hinojosa, C., Ghanem, B.: OpenTAD: A unified framework and comprehensive study of temporal action detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. pp. 2650–2660 (2025)

    2025

  28. [28]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Liu, X., Shi, H., Chen, H., Yu, Z., Li, X., Zhao, G.: iMiGUE: An identity-free video dataset for micro-gesture understanding and emotion analysis. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10631–10642 (2021)

    2021

  29. [29]

    In: International Conference on Learning Representations (2019)

    Loshchilov, I., Hutter, F.: Decoupled weight decay regularization. In: International Conference on Learning Representations (2019)

    2019

  30. [30]

    In: Proceedings of the IJCAI 2025 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2025) (2025)

    Meng, C., Ma, F., Zhang, C., Miao, J., Yang, Y., Zhuang, Y.: Online micro-gesture recognition in long videos via spatiotemporal feature encoding and query-based temporal detection. In: Proceedings of the IJCAI 2025 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2025) (2025)

    2025

  31. [31]

    MiGA Organizers: Miga: Micro-gesture analysis for hidden emotion understanding challenge.https://cv-ac.github.io/MiGA2/(2024), accessed: 2026-05-10

    2024

  32. [32]

    IEEE Transactions on Affective Computing12, 505–523 (2021).https://doi.org/10.1109/taffc.2018.2874986

    Noroozi, F., Corneanu, C.A., Kaminska, D., Sapinski, T., Escalera, S., Anbarjafari, G.: Survey on emotional body gesture recognition. IEEE Transactions on Affective Computing12, 505–523 (2021).https://doi.org/10.1109/taffc.2018.2874986

    work page doi:10.1109/taffc.2018.2874986 2021

  33. [33]

    Journal of Nonverbal Behavior48, 137–159 (2024)

    Poppe, R., van der Zee, S., Taylor, P.J., Anderson, R.J., Veltkamp, R.C.: Mining bodily cues to deception. Journal of Nonverbal Behavior48, 137–159 (2024). https://doi.org/10.1007/s10919-023-00450-9 RGB-Skeleton Fusion for Micro-Gesture Recognition 13

    work page doi:10.1007/s10919-023-00450-9 2024

  34. [34]

    In: Chinese Conference on Pattern Recognition and Computer Vision (PRCV)

    Shang, T., Hao, Y., Pei, M., Li, K., Ben, H., Wang, S.: Cross-modal feature enhancement and contrastive alignment for micro-gesture recognition. In: Chinese Conference on Pattern Recognition and Computer Vision (PRCV). pp. 203–217. Springer (2025)

    2025

  35. [35]

    arXiv preprint arXiv:2606.07355 (2026)

    Shen, X., Li, K., Wang, F., Qian, W., Jiang, J., Guo, D.: Spatial-temporal decoupled adapter for micro-gesture online recognition. arXiv preprint arXiv:2606.07355 (2026)

    Pith/arXiv arXiv 2026

  36. [36]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Shi, D., Zhong, Y., Cao, Q., Ma, L., Li, J., Tao, D.: Tridet: Temporal action detection with relative boundary modeling. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 18857–18866 (2023)

    2023

  37. [37]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Shi, L., Zhang, Y., Cheng, J., Lu, H.: Two-stream adaptive graph convolutional networks for skeleton-based action recognition. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 12026–12035 (2019)

    2019

  38. [38]

    In: Advances in Neural Information Processing Systems

    Tan, J., Zhao, X., Shi, X., Kang, B., Wang, L.: PointTAD: Multi-label temporal action detection with learnable query points. In: Advances in Neural Information Processing Systems. vol. 35 (2022)

    2022

  39. [39]

    In: Advances in Neural Information Processing Systems

    Tong, Z., Song, Y., Wang, J., Wang, L.: VideoMAE: Masked autoencoders are data-efficient learners for self-supervised video pre-training. In: Advances in Neural Information Processing Systems. vol. 35, pp. 10078–10093 (2022)

    2022

  40. [40]

    In: Proceedings of the IJCAI 2024 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2024)

    Wang, Y., Linghu, K., Huang, H., Xia, Z.: Micro-gesture online recognition with dual-stream multi-scale transformer in long videos. In: Proceedings of the IJCAI 2024 Workshop on Micro-gesture Analysis for Hidden Emotion Understanding (MiGA 2024). CEUR Workshop Proceedings, vol. 3848 (2024)

    2024

  41. [41]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Xu, M., Zhao, C., Rojas, D.S., Thabet, A., Ghanem, B.: G-tad: Sub-graph localiza- tion for temporal action detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10156–10165 (2020)

    2020

  42. [42]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Yan, S., Xiong, Y., Lin, D.: Spatial temporal graph convolutional networks for skeleton-based action recognition. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 32 (2018)

    2018

  43. [43]

    In: Proceedings of the European Conference on Computer Vision (2024)

    Yang, L., Zheng, Z., Han, Y., Cheng, H., Song, S., Huang, G., Li, F.: DyFADet: Dynamic feature aggregation for temporal action detection. In: Proceedings of the European Conference on Computer Vision (2024)

    2024

  44. [44]

    In: Proceedings of the European Conference on Computer Vision

    Zhang, C.L., Wu, J., Li, Y.: Actionformer: Localizing moments of actions with transformers. In: Proceedings of the European Conference on Computer Vision. pp. 492–510. Springer (2022)

    2022

  45. [45]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Zheng, Z., Wang, P., Liu, W., Li, J., Ye, R., Ren, D.: Distance-IoU loss: Faster and better learning for bounding box regression. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 34, pp. 12993–13000 (2020)

    2020