Pith. sign in

REVIEW 2 major objections 5 minor 147 references

EquiFusion: Kinematics-Agnostic Human Motion Prediction via Equivariant Latent Diffusion

T0 review · 2 major / 5 minor · reviewed 2026-07-14 · grok-4.5

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

Pith's one-line read One equivariant network predicts future human motion for any skeleton layout, including ones never seen in training.

desk verdict Clean first kinematics-agnostic SHMP via end-to-end joint-order equivariant latent diffusion; zero-shot full/partial skeletons and multi-dataset training work, with real efficiency gains and solid proofs. read the letter →

arxiv 2607.10984 v1 pith:G3TBDHBU submitted 2026-07-13 cs.CV cs.AIcs.HCcs.LG

Cecilia Curreli , Florian Hofherr , Dominik Muhle , Abhishek Saroha , Riccardo Marin , Daniel Cremers This is my paper
classification cs.CVcs.AIcs.HCcs.LG
keywords stochastichumanmotionpredictionkinematics-agnosticpermutationequivariancelatentdiffusionzero-shotkinematicscross-datasettrainingbonedirectionsocclusionhandling
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

Stochastic human motion prediction has long been locked to a single skeleton: every new joint set or capture system forced a fresh network, retargeting errors, and no way to train across datasets. EquiFusion removes that lock. It treats the skeleton’s connectivity matrix as an ordinary input and builds every layer so that permuting the joint order simply permutes the output. The resulting latent diffusion model can therefore be trained once on one or several skeletons and still produce diverse, realistic futures for entirely new full-body layouts, for skeletons with missing limbs, and for out-of-distribution motions. On standard benchmarks it matches or beats the previous best specialised models while using up to 75 percent fewer parameters and running faster. The practical payoff is a single, compact predictor that works across capture systems and occlusion patterns without engineering a new architecture each time.

What carries the argument

End-to-end joint-order permutation equivariance (Lemma 1 / Theorem 2 / Eq. 2): every graph convolution and attention layer is written so that reordering the joints and the adjacency matrix simply reorders the output, guaranteeing that the number of parameters is independent of the number of joints.

What would settle it

Train only on AMASS (22 joints), then evaluate zero-shot on H36M (17 joints) and on randomly occluded AMASS limbs; if the unified ADE, FID and body-realism gaps versus the best retargeted specialised baselines reverse or vanish, the claim fails.

Watch

Extended reading notes

Core claim

Permutation equivariance with respect to joint order, together with an explicit adjacency-matrix input, is sufficient to make a latent diffusion model kinematics-agnostic: a single set of weights can train on heterogeneous skeletons and generalise zero-shot to unseen full-body and partial kinematics while remaining competitive on conventional single-skeleton benchmarks.

Load-bearing premise

That making the network ignore joint order and take connectivity as input is enough for true generalisation across skeletons that have different end-effectors and different numbers of joints, without any joint-type labels or other extra bias.

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

2 major / 5 minor

Summary. The paper proposes EquiFusion, claimed as the first kinematics-agnostic model for stochastic 3D human motion prediction (SHMP). It implements a latent diffusion model whose autoencoder and denoiser are end-to-end permutation-equivariant with respect to joint ordering (Eq. 2), taking the skeleton adjacency matrix A as an explicit input. This design satisfies the parameter-cardinality independence of Lemma 1 via Theorem 2, enabling a single model to train on heterogeneous kinematics (AMASS + Nymeria) and to perform zero-shot inference on unseen full-body skeletons (H36M) and partial/occluded limbs without retargeting or ad-hoc occlusion training. Empirically it reports state-of-the-art or competitive precision, diversity and realism on AMASS, H36M and Nymeria while using up to 75 % fewer parameters than the strongest kinematics-specific diffusion baseline and supporting constant scaling with the number of kinematics.

Significance. If the claims hold, the work removes a long-standing structural bottleneck in SHMP: hard-coded skeleton graphs that force per-dataset models, expensive retargeting, and inability to handle occlusions or novel sensors. The combination of a clean linear-algebra equivariance argument (Secs. C.1–C.2), an explicit adjacency input, bone-direction motion representation that guarantees limb-length consistency by construction, and extensive zero-shot and multi-dataset experiments constitutes a genuine advance toward foundation-style human-motion models. Concrete strengths include machine-checkable equivariance proofs, public code, constant-parameter scaling (Fig. 3), and ablations that isolate the contribution of equivariance (Tab. 4). These results are immediately useful for robotics, AR and multi-sensor pipelines that must operate across heterogeneous MoCap formats.

major comments (2)
  1. Sec. 3.2–3.3 and Lemma 1 / Theorem 2 correctly establish that joint-order permutation equivariance plus an explicit adjacency matrix yields cardinality-independent parameters. The residual claim that this is already sufficient for true non-homeomorphic kinematics-agnosticism (different end-effectors, different joint counts) without joint-type semantics rests primarily on the empirical zero-shot tables (Tabs. 1, 3, 11) and the “Ours+S” ablation. While the ablation shows no gain from semantics, a short additional experiment that systematically varies end-effector sets (or reports failure cases on more distant graphs) would strengthen the sufficiency argument beyond the current human-only skeletons.
  2. Sec. 5 Limitations correctly notes that bone-direction decoding handles only leaf missing joints. Because partial-kinematics zero-shot is advertised as a core novel capability (Abstract, Sec. 1, Tab. 3), the paper should either (a) quantify the fraction of real-world occlusions that are leaf-only or (b) supply a simple non-leaf completion baseline so that readers can judge the practical scope of the claimed occlusion handling.
minor comments (5)
  1. Fig. 3 (left) would be clearer if the y-axis were log-scaled or if absolute parameter counts for each baseline under multi-kinematics were listed in a small table; the visual claim of constant scaling is important and currently a bit hard to read.
  2. Unified metrics (uADE etc.) are a welcome contribution (Sec. 4.1, F.1). A one-sentence reminder in every table caption that conventional ADE ranks identically would help readers who still compare against older papers that report only ADE.
  3. Sec. D.3 and Fig. 5: the forward-kinematics rescaling step that restores bone lengths from the observation is described clearly, yet a short pseudocode block would make the pipeline fully reproducible without consulting the (promised) code.
  4. Occasional typographical inconsistencies appear (e.g., “i.e.kinematics”, missing spaces after commas in several places). A final proof-reading pass would polish the manuscript.
  5. Tab. 4 “Ours on P+Pϵ” versus “Ours on P”: the distinction between sample-wise and distributional equivariance is correctly explained in Sec. C.3; a single clarifying sentence in the table caption would prevent mis-reading by non-specialists.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: equivariance is a pure linear-algebra property independent of data fits, and all empirical claims are evaluated on held-out public splits.

full rationale

Lemma 1 and Theorem 2 (Sec. 3.2) establish that joint-order permutation equivariance implies parameter count independent of J; the proofs in App. C.1–C.2 are ordinary matrix algebra (PW = WP forces shared scalars α, β; graph-conv and attention layers preserve P by construction when A is an explicit input). These derivations do not depend on any fitted quantity, self-cited uniqueness theorem, or ansatz smuggled from prior work. The architecture (Eqs. 3–4) is therefore kinematics-agnostic by design, not by construction from the target metrics. Empirical tables (Tabs. 1–5, 11, 16–18) report standard SHMP metrics on held-out test splits of AMASS/H36M/Nymeria/MoYoga; no parameter is fitted on a subset and then “predicted” on a closely related quantity. Self-citations are limited to the authors’ earlier SkelDiff baseline, which is treated strictly as a competitor (ablated, retargeted, and outperformed), never as an axiom that forces the present claims. Ablations (Tab. 4: “Ours w/o Eq”, multi-dataset SkelDiff variant) further falsify rather than tautologically confirm the design. Consequently the central “first kinematics-agnostic SHMP” claim rests on independent mathematical sufficiency plus ordinary held-out evaluation; no circular reduction exists.

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

The central claim rests on standard linear-algebra facts about permutation equivariance, the domain modeling of skeletons as undirected graphs, and a small set of architectural hyper-parameters chosen once and held fixed across all experiments. No new physical entities are postulated; the free parameters are ordinary neural-network design choices.

free parameters (3)
  • latent dimension L = 96
    Fixed at 96 (4× temporal compression) following prior SHMP latent models; not fitted to the zero-shot metrics.
  • diffusion steps T and cosine schedule = T=10
    T=10 chosen for speed; schedule is standard. Same values used for all datasets.
  • training epochs / learning rate / k-relaxation
    300 AE epochs, 375 denoiser epochs, lr=0.005, k=50 for diversity; held constant across experiments.
assumptions (3)
  • standard math A network operation o(X)=WXG is permutation-equivariant under joint reordering if and only if its weight matrix satisfies PW=WP, which forces the parameter count to be independent of joint cardinality J (Lemma 1 / Theorem 2).
    Standard linear-algebra fact used to justify the architecture; proved in Sec. C.1.
  • domain assumption Human skeletons can be modeled as undirected graphs whose adjacency matrix fully captures the kinematic connectivity needed for motion prediction.
    Standard in GCN-based HMP; invoked throughout Sec. 3.
  • domain assumption Bone-direction vectors (relative parent-to-child) plus rescaling by observed limb lengths guarantee constant bone lengths without singularities.
    Used for the motion parametrization in Sec. 3.4; empirically validated but not mathematically necessary for equivariance.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EquiFusion: Kinematics-Agnostic Human Motion Prediction via Equivariant Latent Diffusion." pith.science (2026). https://pith.science/paper/G3TBDHBU

@misc{pith2026260710984,
  author       = {Pith},
  title        = {Pith review of: EquiFusion: Kinematics-Agnostic Human Motion Prediction via Equivariant Latent Diffusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G3TBDHBU}},
  note         = {Machine review of arXiv:2607.10984}
}
read the original abstract

Existing Stochastic 3D Human Motion Prediction models are fundamentally constrained by hard-coding the skeleton kinematics, severely limiting generalization, preventing cross-dataset training, and requiring complex data retargeting. We introduce EquiFusion, the first kinematics-agnostic model to solve this bottleneck, implementing a latent diffusion model with a permutation equivariant architecture. EquiFusion treats the kinematics' connectivity as an explicit input parameter, ensuring its internal computations are inherently agnostic to joint ordering and graph structure. This novel design enables truly cross-dataset generalization to unseen kinematics and unlocks novel zero-shot directions, such as motion prediction from partial or occluded observations and targeted limb generation. EquiFusion achieves state-of-the-art results on major benchmarks, being up to 75% more compact than previous kinematics-specific methods, while achieving faster training and inference. EquiFusion thus establishes a new, flexible standard for robust human motion prediction. Model and training code are available at https://ceveloper.github.io/publications/equifusion/.

Figures

Figures reproduced from arXiv: 2607.10984 by the authors.

Figure 1
Figure 1. EquiFusion. We introduce the first model for stochastic human motion pre￾diction that generalizes to unseen skeleton parameterization, i.e. kinematics. While previous methods require a trained instance for each dataset or better kinematics, with a single model we unlock training on multiple datasets and inference on motion parametrized with different kinematics. EquiFusion is the first SHMP model to handle zero-shot… view at source ↗
Figure 2
Figure 2. EquiFusion is the first skeleton-agnostic model for SHMP, imple [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. State-of-the-art with 75% fewer parameters [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Qualitatives for zero-shot kinematics on H36M( [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Pipeline overview of parametrizing motions as bone directions [PITH_FULL_IMAGE:figures/full_fig_p042_5.png]
Figure 6
Figure 6. Figure 6: Overview of the architecture layers of our model. [PITH_FULL_IMAGE:figures/full_fig_p043_6.png]
Figure 7
Figure 7. Figure 7: Qualitative Results for the cross-topology experiment on H36M of Tab. 1. We report out of 50 predictions, the sample closest to the GT, and the two predictions that maximize diversity when paired with the closest to GT sample. Segment n. 605. Past Future Closest to GT …
Figure 8
Figure 8. Figure 8: Qualitative Results for the cross-topology experiment on H36M of Tab. 1. Segment n. 1774 [PITH_FULL_IMAGE:figures/full_fig_p060_8.png]
Figure 9
Figure 9. Figure 9: Qualitative Results for the out-of-distribution testing on the MoCap Yoga dataset Tab. 2. Segment n. 4651. As the setting is quite challenging, we report only the example closest to GT. Past Future Closest to GT CoMusion SkeletonDiffusion Ours Self penetration Center o…
Figure 10
Figure 10. Figure 10: Qualitative Results for the out-of-distribution testing on the MoCap Yoga dataset Tab. 2. Segment n. 4656 [PITH_FULL_IMAGE:figures/full_fig_p061_10.png]
Figure 11
Figure 11. Figure 11: Qualitative Example of missing left arm in the observation. SkelDiff has been paired with the symmetric limb pipeline for input completion. Test on AMASS Segment n. 12324. Past Future Closest to GT SkeletonDiffusion Ours ...static... Dynamics on left leg [PITH_FULL_I…
Figure 12
Figure 12. Figure 12: Qualitative Example of missing both arms in the observation. SkelDiff can only be paired with the restpose approach to complete the input before further processing. Test on AMASS Segment n. 11100 [PITH_FULL_IMAGE:figures/full_fig_p062_12.png]
Figure 13
Figure 13. Figure 13: Qualitative Example of missing both arms in the observation in a cross￾topology setting. We do not compare with other methods, as they would require being extended with both retargeting and completion and be exposed to too high degradation. Test on H36M Segment n. 200…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

147 extracted references · 17 linked inside Pith

  1. [1]

    Movella XSens MVN Link motion capture,https : / / www . movella . com / products/motion-capture/xsens-mvn-link3

  2. [2]

    ACM Transactions on Graphics (TOG)39(4), 62–1 (2020) 4, 2

    Aberman, K., Li, P., Lischinski, D., Sorkine-Hornung, O., Cohen-Or, D., Chen, B.: Skeleton-aware networks for deep motion retargeting. ACM Transactions on Graphics (TOG)39(4), 62–1 (2020) 4, 2

    2020

  3. [3]

    arXiv preprint arXiv:1905.01680 (2019) 4, 2

    Aberman, K., Wu, R., Lischinski, D., Chen, B., Cohen-Or, D.: Learning character- agnostic motion for motion retargeting in 2d. arXiv preprint arXiv:1905.01680 (2019) 4, 2

    arXiv 1905

  4. [4]

    Nature630(8016), 493–500 (2024) 9

    Abramson, J., Adler, J., Dunger, J., Evans, R., Green, T., Pritzel, A., Ron- neberger, O., Willmore, L., Ballard, A.J., Bambrick, J., et al.: Accurate struc- ture prediction of biomolecular interactions with alphafold 3. Nature630(8016), 493–500 (2024) 9

    2024

  5. [5]

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

    Adeli, V., Ehsanpour, M., Reid, I., Niebles, J.C., Savarese, S., Adeli, E., Rezatofighi, H.: Tripod: Human trajectory and pose dynamics forecasting in the wild. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 13390–13400 (2021) 3

    2021

  6. [6]

    In: 2021 International Conference on 3D Vision (3DV)

    Aksan, E., Kaufmann, M., Cao, P., Hilliges, O.: A spatio-temporal transformer for 3d human motion prediction. In: 2021 International Conference on 3D Vision (3DV). pp. 565–574. IEEE (2021) 3

    2021

  7. [7]

    Frontiers in Robotics and AI9, 809114 (2022) 7

    AlAttar, A., Chappell, D., Kormushev, P.: Kinematic-model-free predictive con- trol for robotic manipulator target reaching with obstacle avoidance. Frontiers in Robotics and AI9, 809114 (2022) 7

    2022

  8. [8]

    IEEE Transactions on Robotics33(1), 153–168 (2016) 2, 4

    Alibeigi, M., Ahmadabadi, M.N., Araabi, B.N.: A fast, robust, and incremental model for learning high-level concepts from human motions by imitation. IEEE Transactions on Robotics33(1), 153–168 (2016) 2, 4

    2016

Show all 147 references
  1. [9]

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

    Barquero, G., Escalera, S., Palmero, C.: Belfusion: Latent diffusion for behavior- driven human motion prediction. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 2317–2327 (2023) 2, 4, 5, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 21, 24, 26, 27...

    2023

  2. [10]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops

    Barsoum, E., Kender, J., Liu, Z.: Hp-gan: Probabilistic 3d human motion pre- diction via gan. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops. pp. 1418–1427 (2018) 4

    2018

  3. [11]

    arXiv preprint arXiv:2204.01565 (2022) 10, 18

    Bie, X., Guo, W., Leglaive, S., Girin, L., Moreno-Noguer, F., Alameda-Pineda, X.: Hit-dvae: Human motion generation via hierarchical transformer dynamical vae. arXiv preprint arXiv:2204.01565 (2022) 10, 18

    2022 arXiv

  4. [12]

    Advances in neural information processing systems34, 18673– 18684 (2021) 3, 9

    Bietti, A., Venturi, L., Bruna, J.: On the sample complexity of learning under geometric stability. Advances in neural information processing systems34, 18673– 18684 (2021) 3, 9

    2021

  5. [13]

    OpenAI Blog1(8), 1 (2024) 2

    Brooks, T., Peebles, B., Holmes, C., DePue, W., Guo, Y., Jing, L., Schnurr, D., Taylor, J., Luhman, T., Luhman, E., et al.: Video generation models as world simulators. OpenAI Blog1(8), 1 (2024) 2

    2024

  6. [14]

    In: Computer Vision–ECCV 2020: 16th European Conference, Glas- gow, UK, August 23–28, 2020, Proceedings, Part VII 16

    Cai, Y., Huang, L., Wang, Y., Cham, T.J., Cai, J., Yuan, J., Liu, J., Yang, X., Zhu, Y., Shen, X., et al.: Learning progressive joint propagation for human motion prediction. In: Computer Vision–ECCV 2020: 16th European Conference, Glas- gow, UK, August 23–28, 2020, Proceeding...

    2020

  7. [15]

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

    Cai, Y., Wang, Y., Zhu, Y., Cham, T.J., Cai, J., Yuan, J., Liu, J., Zheng, C., Yan, S., Ding, H., et al.: A unified 3d human motion synthesis model via condi- EquiFusion: Kinematics-Agnostic HMP 17 tional variational auto-encoder. In: Proceedings of the IEEE/CVF International ...

    2021

  8. [16]

    In: CVPR (2017) 5

    Cao, Z., Simon, T., Wei, S.E., Sheikh, Y.: Realtime multi-person 2d pose estima- tion using part affinity fields. In: CVPR (2017) 5

    2017

  9. [17]

    arXiv preprint arXiv:2505.20857 (2025) 4, 5, 2

    Cao, Z., Liu, B., Li, S., Zhang, W., Chen, H.: G-dream: Graph-conditioned dif- fusion retargeting across multiple embodiments. arXiv preprint arXiv:2505.20857 (2025) 4, 5, 2

    2025 arXiv

  10. [18]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Chen, L.H., Zhang, J., Li, Y., Pang, Y., Xia, X., Liu, T.: Humanmac: Masked mo- tion completion for human motion prediction. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 9544–9555 (2023) 2, 4, 7, 9, 10, 11, 14, 16, 22, 24, 26, 30, 31, 32

    2023

  11. [19]

    The Journal of Visualization and Computer Animation11(5), 223–235 (2000) 4, 2

    Choi, K.J., Ko, H.S.: Online motion retargetting. The Journal of Visualization and Computer Animation11(5), 223–235 (2000) 4, 2

    2000

  12. [20]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Ci,H.,Wu,M.,Zhu,W.,Ma,X.,Dong,H.,Zhong,F.,Wang,Y.:Gfpose:Learning 3d human pose prior with gradient fields. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 4800–4810 (2023) 6

    2023

  13. [21]

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

    Cui, Q., Sun, H.: Towards accurate 3d human motion prediction from incomplete observations. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 4801–4810 (2021) 2, 4, 5

    2021

  14. [22]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Cui, Q., Sun, H., Yang, F.: Learning dynamic relationships for 3d human motion prediction. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 6519–6527 (2020) 4, 3

    2020

  15. [23]

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

    Curreli, C., Muhle, D., Saroha, A., Ye, Z., Marin, R., Cremers, D.: Nonisotropic gaussian diffusion for realistic 3d human motion prediction. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1871–1882 (2025) 2, 3, 4, 5, 7, 8, 9, 10, 11...

    2025

  16. [24]

    Dabhi,M.,Jeni,L.A.,Lucey,S.:3d-lfm:Liftingfoundationmodel.In:Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10466–10475 (2024) 4, 6

    2024

  17. [25]

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

    Dang, L., Nie, Y., Long, C., Zhang, Q., Li, G.: Msr-gcn: Multi-scale residual graph convolution networks for human motion prediction. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 11467–11476 (2021) 3

    2021

  18. [26]

    In: Proceedings of the 30th ACM International Conference on Multimedia

    Dang, L., Nie, Y., Long, C., Zhang, Q., Li, G.: Diverse human motion prediction via gumbel-softmax sampling from an auxiliary space. In: Proceedings of the 30th ACM International Conference on Multimedia. pp. 5162–5171 (2022) 2, 4, 5, 7, 9, 10, 11, 12, 8, 15, 16, 24, 26, 27, 2...

    2022

  19. [27]

    In: 2017 Inter- national Joint Conference on Neural Networks (IJCNN)

    Delhaisse, B., Esteban, D., Rozo, L., Caldwell, D.: Transfer learning of shared latent spaces between robots with similar kinematic structure. In: 2017 Inter- national Joint Conference on Neural Networks (IJCNN). pp. 4142–4149. IEEE (2017) 4, 2

    2017

  20. [28]

    Advances in neural information processing systems34, 8780–8794 (2021) 8

    Dhariwal, P., Nichol, A.: Diffusion models beat gans on image synthesis. Advances in neural information processing systems34, 8780–8794 (2021) 8

    2021

  21. [29]

    arXiv preprint arXiv:2308.13561 (2023) 3

    Engel, J., Somasundaram, K., Goesele, M., Sun, A., Gamino, A., Turner, A., Talattof, A., Yuan, A., Souti, B., Meredith, B., et al.: Project aria: A new tool for egocentric multi-modal ai research. arXiv preprint arXiv:2308.13561 (2023) 3

    2023 arXiv

  22. [30]

    In: Motion in Games: 5th International Conference, MIG 2012, Rennes, France, November 15-17, 2012

    Feng, A., Huang, Y., Xu, Y., Shapiro, A.: Automating the transfer of a generic set of behaviors onto a virtual character. In: Motion in Games: 5th International Conference, MIG 2012, Rennes, France, November 15-17, 2012. Proceedings 5. pp. 134–145. Springer (2012) 4, 2 18 C. C...

    2012

  23. [31]

    In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    Figuera, X., Park, S., Ahn, H.: Redefining data pairing for motion retargeting leveraging a human body prior. In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). pp. 4800–4807. IEEE (2024) 4

    2024

  24. [32]

    In: Proceedings of the IEEE international conference on com- puter vision

    Fragkiadaki, K., Levine, S., Felsen, P., Malik, J.: Recurrent network models for human dynamics. In: Proceedings of the IEEE international conference on com- puter vision. pp. 4346–4354 (2015) 3

    2015

  25. [33]

    Authorea Preprints (2025) 3

    Fritsche, O., Camacho, S., Hossain, M.S.B., Halpenny, T., Archniegas, C., Dranetz, J., Hadley, D., Guo, Z., Choi, H.: Ultra-mocap: A multimodal imu and semg dataset for upper body joint kinematics analysis. Authorea Preprints (2025) 3

    2025

  26. [34]

    In: Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers

    Gat, I., Raab, S., Tevet, G., Reshef, Y., Bermano, A.H., Cohen-Or, D.: Anytop: Character animation diffusion with any topology. In: Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers. pp. 1–10 (2025) 4, 7, 6

    2025

  27. [35]

    arXiv preprint arXiv:2404.11735 (2024) 5, 10, 18

    Geist, A.R., Frey, J., Zhobro, M., Levina, A., Martius, G.: Learning with 3d rotations, a hitchhiker’s guide to so (3). arXiv preprint arXiv:2404.11735 (2024) 5, 10, 18

    2024 arXiv

  28. [36]

    In: Iberian Robotics conference

    Gil, O., Sanfeliu, A.: Human motion trajectory prediction using the social force model for real-time and low computational cost applications. In: Iberian Robotics conference. pp. 235–247. Springer (2023) 7

    2023

  29. [37]

    In: Proceedings of the 25th annual conference on Computer graphics and interactive techniques

    Gleicher, M.: Retargetting motion to new characters. In: Proceedings of the 25th annual conference on Computer graphics and interactive techniques. pp. 33–42 (1998) 4, 2

    1998

  30. [38]

    Pattern Recognition146, 109998 (2024) 4

    Gu, C., Yu, J., Zhang, C.: Learning disentangled representations for controllable human motion prediction. Pattern Recognition146, 109998 (2024) 4

    2024

  31. [39]

    In: Proceedings of the european conference on computer vision (ECCV)

    Gui, L.Y., Wang, Y.X., Liang, X., Moura, J.M.: Adversarial geometry-aware hu- man motion prediction. In: Proceedings of the european conference on computer vision (ECCV). pp. 786–803 (2018) 3

    2018

  32. [40]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Gupta, A., Johnson, J., Fei-Fei, L., Savarese, S., Alahi, A.: Social gan: Socially acceptable trajectories with generative adversarial networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2255–2264 (2018) 17

    2018

  33. [41]

    In: 2024 International Conference on 3D Vision (3DV)

    Guzov, V., Chibane, J., Marin, R., He, Y., Saracoglu, Y., Sattler, T., Pons-Moll, G.: Interaction replica: Tracking human–object interaction and scene changes from human motion. In: 2024 International Conference on 3D Vision (3DV). pp. 1006–1016. IEEE (2024) 3

    2024

  34. [42]

    Advances in neural information processing systems33, 6840–6851 (2020) 2, 8, 13, 17, 20

    Ho, J., Jain, A., Abbeel, P.: Denoising diffusion probabilistic models. Advances in neural information processing systems33, 6840–6851 (2020) 2, 8, 13, 17, 20

    2020

  35. [43]

    ACM Transactions on Graphics (TOG)35(4), 1–11 (2016) 2, 4, 11, 12, 22, 26, 27, 28, 32

    Holden,D.,Saito,J.,Komura,T.:Adeeplearningframeworkforcharactermotion synthesis and editing. ACM Transactions on Graphics (TOG)35(4), 1–11 (2016) 2, 4, 11, 12, 22, 26, 27, 28, 32

    2016

  36. [44]

    IEEE Transactions on Visualization and Computer Graphics (2023) 4, 7, 2

    Hu, L., Zhang, Z., Zhong, C., Jiang, B., Xia, S.: Pose-aware attention network for flexible motion retargeting by body part. IEEE Transactions on Visualization and Computer Graphics (2023) 4, 7, 2

    2023

  37. [45]

    In: Proceedings of the SIGGRAPH Asia 2025 Conference Papers

    Huang, Z., Feng, H., Sun, Y.T., Guo, Y.C., Cao, Y.P., Sheng, L.: Animax: Ani- mating the inanimate in3d with joint video-pose diffusion models. In: Proceedings of the SIGGRAPH Asia 2025 Conference Papers. pp. 1–13 (2025) 4

    2025

  38. [46]

    Ionescu, C., Papava, D., Olaru, V., Sminchisescu, C.: Human3.6m: Large scale datasets and predictive methods for 3d human sensing in natural environments (2014),http://vision.imar.ro/human3.6m.2, 3, 6, 9, 10, 11, 12, 14, 26, 27, 30, 31, 32 EquiFusion: Kinematics-Agnostic HMP 19

    2014

  39. [47]

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

    Jain, A., Zamir, A.R., Savarese, S., Saxena, A.: Structural-rnn: Deep learning on spatio-temporal graphs. In: Proceedings of the ieee conference on computer vision and pattern recognition. pp. 5308–5317 (2016) 3

    2016

  40. [48]

    In: SIGGRAPH Asia 2018 Posters

    Jang, H., Kwon, B., Yu, M., Kim, S.U., Kim, J.: A variational u-net for motion retargeting. In: SIGGRAPH Asia 2018 Posters. pp. 1–2 (2018) 4, 2

    2018

  41. [49]

    ACM Transactions on Graphics (TOG)43(6), 1–17 (2024) 4, 5, 2

    Jang, I., Choi, S., Hong, S., Kim, C., Noh, J.: Geometry-aware retargeting for two-skinned characters interaction. ACM Transactions on Graphics (TOG)43(6), 1–17 (2024) 4, 5, 2

    2024

  42. [50]

    Advances in neural information processing systems32(2019) 9

    Keriven,N.,Peyré,G.:Universalinvariantandequivariantgraphneuralnetworks. Advances in neural information processing systems32(2019) 9

    2019

  43. [51]

    arXiv preprint arXiv:2509.25600 (2025) 4, 2

    Kim, W., Li, T., Ha, S.: Moreflow: Motion retargeting learning through unsuper- vised flow matching. arXiv preprint arXiv:2509.25600 (2025) 4, 2

    2025

  44. [52]

    arXiv preprint arXiv:1609.02907 (2016) 9, 4

    Kipf, T.N., Welling, M.: Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907 (2016) 9, 4

    2016 arXiv

  45. [53]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Kolotouros, N., Pavlakos, G., Black, M.J., Daniilidis, K.: Learning to reconstruct 3d human pose and shape via model-fitting in the loop. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 2252–2261 (2019) 3

    2019

  46. [54]

    In: Proceedings of the AAAI conference on artificial intelligence

    Kundu, J.N., Gor, M., Babu, R.V.: Bihmp-gan: Bidirectional 3d human motion prediction gan. In: Proceedings of the AAAI conference on artificial intelligence. vol. 33, pp. 8553–8560 (2019) 4

    2019

  47. [55]

    In: International Confer- ence on Learning Representations

    Laabid, N., Rissanen, S., Heinonen, M., Solin, A., Garg, V.: Equivariant denoisers cannot copy graphs: Align your graph diffusion models. In: International Confer- ence on Learning Representations. vol. 2025, pp. 100256–100294 (2025) 4

    2025

  48. [56]

    In: Proceedings of the 26th annual conference on Computer graphics and interactive techniques

    Lee, J., Shin, S.Y.: A hierarchical approach to interactive motion editing for human-like figures. In: Proceedings of the 26th annual conference on Computer graphics and interactive techniques. pp. 39–48 (1999) 4, 2

    1999

  49. [57]

    Lee, S., Kang, T., Park, J., Lee, J., Won, J.: Same: Skeleton-agnostic motion embeddingforcharacteranimation.In:SIGGRAPHAsia2023ConferencePapers. pp. 1–11 (2023) 2, 4, 7, 6

    2023

  50. [58]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Li, C., Zhang, Z., Lee, W.S., Lee, G.H.: Convolutional sequence to sequence model for human dynamics. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 5226–5234 (2018) 3

    2018

  51. [59]

    In: 2025 International Conference on 3D Vision (3DV)

    Li, C., Chibane, J., He, Y., Pearl, N., Geiger, A., Pons-Moll, G.: Unimotion: Unifying 3d human motion synthesis and understanding. In: 2025 International Conference on 3D Vision (3DV). pp. 240–249. IEEE (2025) 5

    2025

  52. [60]

    In: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition

    Li, M., Chen, S., Chen, X., Zhang, Y., Wang, Y., Tian, Q.: Actional-structural graph convolutional networks for skeleton-based action recognition. In: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 3595–3603 (2019) 4, 3

    2019

  53. [61]

    In: Proceed- ings of the IEEE/CVF international conference on computer vision

    Li, M., Chen, S., Liu, Z., Zhang, Z., Xie, L., Tian, Q., Zhang, Y.: Skeleton graph scattering networks for 3d skeleton-based human motion prediction. In: Proceed- ings of the IEEE/CVF international conference on computer vision. pp. 854–864 (2021) 9, 3

    2021

  54. [62]

    In: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recogni- tion

    Li, M., Chen, S., Zhao, Y., Zhang, Y., Wang, Y., Tian, Q.: Dynamic multiscale graph neural networks for 3d skeleton based human motion prediction. In: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recogni- tion. pp. 214–223 (2020) 4, 7, 9, 3

    2020

  55. [63]

    In: ACM SIGGRAPH 2024 Conference Papers

    Li, P., Starke, S., Ye, Y., Sorkine-Hornung, O.: Walkthedog: Cross-morphology motion alignment via phase manifolds. In: ACM SIGGRAPH 2024 Conference Papers. pp. 1–10 (2024) 4, 2 20 C. Curreli et al

    2024

  56. [64]

    In: SIGGRAPH Asia 2023 Conference Papers

    Li, T., Won, J., Clegg, A., Kim, J., Rai, A., Ha, S.: Ace: Adversarial correspon- dence embedding for cross morphology motion retargeting from human to nonhu- man characters. In: SIGGRAPH Asia 2023 Conference Papers. pp. 1–11 (2023) 4

    2023

  57. [65]

    In: European Conference on Computer Vision

    Liao, Z., Yang, J., Saito, J., Pons-Moll, G., Zhou, Y.: Skeleton-free pose transfer for stylized 3d characters. In: European Conference on Computer Vision. pp. 640–656. Springer (2022) 4, 5, 2

    2022

  58. [66]

    In: 30th British Machine Vision Con- ference (BMVC 2019)

    Lim, J., Chang, H.J., Choi, J.Y.: Pmnet: Learning of disentangled pose and move- ment for unsupervised motion retargeting. In: 30th British Machine Vision Con- ference (BMVC 2019). British Machine Vision Association, BMVA (2019) 2, 4

    2019

  59. [67]

    arXiv preprint arXiv:2512.07289 (2025) 8, 13, 14

    Lin, P., Chen, P., Jiao, R., Mo, Q., Cen, J., Huang, W., Liu, Y., Huang, D., Lu, Y.: Equivariant diffusion for crystal structure prediction. arXiv preprint arXiv:2512.07289 (2025) 8, 13, 14

    2025

  60. [68]

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

    Liu, Q., Lv, K., Dong, K., Xue, J., Niu, Z., Wang, J.: Text-to-any-skeleton motion generation without retargeting. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 12926–12936 (2025) 4, 7, 6

    2025

  61. [69]

    In: Proceedings of the AAAI conference on artificial intelligence

    Liu, Z., Lyu, K., Wu, S., Chen, H., Hao, Y., Ji, S.: Aggregated multi-gans for controlled 3d human motion prediction. In: Proceedings of the AAAI conference on artificial intelligence. vol. 35, pp. 2225–2232 (2021) 4

    2021

  62. [70]

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

    Liu,Z.,Wu,S.,Jin,S.,Liu,Q.,Lu,S.,Zimmermann,R.,Cheng,L.:Towardsnatu- ral and accurate future motion prediction of humans and animals. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10004–10012 (2019) 3

    2019

  63. [71]

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

    Lohit, S., Anirudh, R., Turaga, P.: Recovering trajectories of unmarked joints in 3d human actions using latent space optimization. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. pp. 2342– 2351 (2021) 5

    2021

  64. [72]

    ACM Trans

    Loper, M., Mahmood, N., Romero, J., Pons-Moll, G., Black, M.J.: SMPL: A skinned multi-person linear model. ACM Trans. Graphics (Proc. SIGGRAPH Asia)34(6), 248:1–248:16 (Oct 2015) 3, 5, 10

    2015

  65. [73]

    arXiv preprint arXiv:2005.00178 (2020) 3, 9

    Lyle, C., van der Wilk, M., Kwiatkowska, M., Gal, Y., Bloem-Reddy, B.: On the benefits of invariance in neural networks. arXiv preprint arXiv:2005.00178 (2020) 3, 9

    2005 arXiv

  66. [74]

    In: Pacific-Asia conference on knowledge discovery and data mining

    Ma, L., Rabbany, R., Romero-Soriano, A.: Graph attention networks with posi- tional embeddings. In: Pacific-Asia conference on knowledge discovery and data mining. pp. 514–527. Springer (2021) 35

    2021

  67. [75]

    In: European Conference on Computer Vision

    Ma, L., Ye, Y., Hong, F., Guzov, V., Jiang, Y., Postyeni, R., Pesqueira, L., Gamino, A., Baiyya, V., Kim, H.J., et al.: Nymeria: A massive collection of multi- modal egocentric daily motion in the wild. In: European Conference on Computer Vision. pp. 445–465. Springer (2024) 2...

    2024

  68. [76]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Mahmood, N., Ghorbani, N., Troje, N.F., Pons-Moll, G., Black, M.J.: Amass: Archive of motion capture as surface shapes. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 5442–5451 (2019) 2, 3, 6, 10, 11, 12, 13, 14, 15, 26, 28, 31

    2019

  69. [77]

    In: Computer Vision–ECCV 2020: 16th European Confer- ence, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16

    Mao, W., Liu, M., Salzmann, M.: History repeats itself: Human motion prediction via motion attention. In: Computer Vision–ECCV 2020: 16th European Confer- ence, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16. pp. 474–489. Springer (2020) 3

    2020

  70. [78]

    In: Proceedings of the IEEE/CVF International Con- EquiFusion: Kinematics-Agnostic HMP 21 ference on Computer Vision

    Mao, W., Liu, M., Salzmann, M.: Generating smooth pose sequences for diverse human motion prediction. In: Proceedings of the IEEE/CVF International Con- EquiFusion: Kinematics-Agnostic HMP 21 ference on Computer Vision. pp. 13309–13318 (2021) 2, 4, 5, 9, 11, 7, 8, 16, 26, 27, ...

    2021

  71. [79]

    In: Proceedings of the IEEE/CVF international con- ference on computer vision

    Mao, W., Liu, M., Salzmann, M., Li, H.: Learning trajectory dependencies for human motion prediction. In: Proceedings of the IEEE/CVF international con- ference on computer vision. pp. 9489–9497 (2019) 7, 3

    2019

  72. [80]

    In: European Conference on Computer Vision (ECCV) (sep 2018) 6

    von Marcard, T., Henschel, R., Black, M., Rosenhahn, B., Pons-Moll, G.: Recov- ering accurate 3d human pose in the wild using imus and a moving camera. In: European Conference on Computer Vision (ECCV) (sep 2018) 6

    2018

  73. [81]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Martinez, J., Black, M.J., Romero, J.: On human motion prediction using recur- rent neural networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2891–2900 (2017) 3

    2017

  74. [82]

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

    Martínez-González, A., Villamizar, M., Odobez, J.M.: Pose transformers (potr): Human motion prediction with non-autoregressive transformers. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 2276–2284 (2021) 3

    2021

  75. [83]

    In: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition

    Medjaouri, O., Desai, K.: Hr-stan: High-resolution spatio-temporal attention net- work for 3d human motion prediction. In: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition. pp. 2540–2549 (2022) 3

    2022

  76. [84]

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

    Mo, C.A., Hu, K., Long, C., Yuan, D., Siu, W.C., Wang, Z.: Pumps: Skeleton- agnostic point-based universal motion pre-training for synthesis in human motion tasks. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 14496–14506 (2025) 6

    2025

  77. [85]

    arXiv preprint arXiv:2305.18897 (2023) 4, 5, 2

    Mourot, L., Hoyet, L., Clerc, F.L., Hellier, P.: Humot: human motion represen- tation using topology-agnostic transformers for character animation retargeting. arXiv preprint arXiv:2305.18897 (2023) 4, 5, 2

    2023 arXiv

  78. [86]

    Proceedings of the ACM on Computer Graphics and Interactive Techniques4(3), 1–15 (2021) 4, 2

    Musoni, P., Marin, R., Melzi, S., Castellani, U.: A functional skeleton transfer. Proceedings of the ACM on Computer Graphics and Interactive Techniques4(3), 1–15 (2021) 4, 2

    2021

  79. [87]

    In: STAG

    Musoni,P.,Marin,R.,Melzi,S.,Castellani,U.:Reposingandretargetingunrigged characters with intrinsic-extrinsic transfer. In: STAG. pp. 21–30 (2021) 4, 2

    2021

  80. [88]

    In: 2025 IEEE In- ternational Conference on Bioinformatics and Biomedicine (BIBM)

    Nan, X., Liu, X., You, X., Du, Y., Ji, C., Song, J.: Pgdiff: A physics-guided equivariant diffusion model for structure-based drug design. In: 2025 IEEE In- ternational Conference on Bioinformatics and Biomedicine (BIBM). pp. 687–692 (2025) 4

    2025

  81. [89]

    In: International Conference on Machine Learning

    Nichol, A.Q., Dhariwal, P.: Improved denoising diffusion probabilistic models. In: International Conference on Machine Learning. pp. 8162–8171. PMLR (2021) 20

    2021

  82. [90]

    arXiv preprint arXiv:2006.00424 (2020) 5

    Park, J.S., Manocha, D.: Hmpo: Human motion prediction in occluded environ- ments for safe motion planning. arXiv preprint arXiv:2006.00424 (2020) 5

    2006 arXiv

  83. [91]

    arXiv preprint arXiv:2104.10350 (2021) 8

    Patterson, D., Gonzalez, J., Le, Q., Liang, C., Munguia, L.M., Rothchild, D., So, D., Texier, M., Dean, J.: Carbon emissions and large neural network training. arXiv preprint arXiv:2104.10350 (2021) 8

    2021 arXiv

  84. [92]

    In: Proceedings IEEE Conf

    Pavlakos, G., Choutas, V., Ghorbani, N., Bolkart, T., Osman, A.A.A., Tzionas, D., Black, M.J.: Expressive body capture: 3d hands, face, and body from a single image. In: Proceedings IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) (2019) 3

    2019

  85. [93]

    arXiv preprint arXiv:1805.06485 (2018) 3

    Pavllo,D.,Grangier,D.,Auli,M.:Quaternet:Aquaternion-basedrecurrentmodel for human motion. arXiv preprint arXiv:1805.06485 (2018) 3

    2018 arXiv

  86. [94]

    In: Proceedings of the European conference on computer vision (ECCV) (2026) 3 22 C

    Petrov, I.A., Guzov, V., Marin, R., Aksan, E., Chen, X., Cremers, D., Beeler, T., Pons-Moll, G.: Echo: Ego-centric modeling of human-object interactions. In: Proceedings of the European conference on computer vision (ECCV) (2026) 3 22 C. Curreli et al

    2026

  87. [95]

    Neurocomputing p

    Phu, K.A., Hoang, V.D., et al.: Predicting occluded skeletal joints via tracking- based feature extraction. Neurocomputing p. 131004 (2025) 5

    2025

  88. [96]

    arXiv preprint arXiv:2204.06125 1(2), 3 (2022) 9, 17

    Ramesh, A., Dhariwal, P., Nichol, A., Chu, C., Chen, M.: Hierarchical text- conditional image generation with clip latents. arXiv preprint arXiv:2204.06125 1(2), 3 (2022) 9, 17

    2022 arXiv

  89. [97]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Rempe, D., Birdal, T., Hertzmann, A., Yang, J., Sridhar, S., Guibas, L.J.: Hu- mor: 3d human motion model for robust pose estimation. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 11488–11499 (2021) 3

    2021

  90. [98]

    Rombach, R., Blattmann, A., Lorenz, D., Esser, P., Ommer, B.: High-resolution image synthesis with latent diffusion models (2022) 2, 8, 16, 17

    2022

  91. [99]

    In: 2008 International Symposium on Telecommunications

    Roudsarabi, N., Behrad, A.R.: Solving occlusion problem in 3d human motion reconstruction. In: 2008 International Symposium on Telecommunications. pp. 701–706. IEEE (2008) 5

    2008

  92. [100]

    Saadatnejad, S., Rasekh, A., Mofayezi, M., Medghalchi, Y., Rajabzadeh, S., Mor- dan, T., Alahi, A.: A generic diffusion-based approach for 3d human pose predic- tion in the wild (2023) 4, 9

    2023

  93. [101]

    arXiv preprint (2026) 5

    Saito, J., Li, J., de Ruyter, M., Guerrero, M., Lim, E., Hassani, E., Ribera, R.B., Moon, H., Dadela, M., Di Lucca, M., et al.: Soma: Unifying parametric human body models. arXiv preprint (2026) 5

    2026

  94. [102]

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

    Salzmann, T., Pavone, M., Ryll, M.: Motron: Multimodal probabilistic human motion forecasting. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 6457–6466 (2022) 5, 7, 9, 10, 4, 18, 20

    2022

  95. [103]

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

    Sárándi, I., Hermans, A., Leibe, B.: Learning 3d human pose estimation from dozensofdatasetsusingageometry-awareautoencodertobridgebetweenskeleton formats. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. pp. 2956–2966 (2023) 4, 5

    2023

  96. [104]

    Advances in Neural Information Processing Systems37, 140032–140065 (2025) 4, 3

    Sárándi, I., Pons-Moll, G.: Neural localizer fields for continuous 3d human pose and shape estimation. Advances in Neural Information Processing Systems37, 140032–140065 (2025) 4, 3

    2025

  97. [105]

    Nature Computational Science4(12), 899–909 (2024) 4

    Schneuing, A., Harris, C., Du, Y., Didi, K., Jamasb, A., Igashov, I., Du, W., Gomes, C., Blundell, T.L., Lio, P., et al.: Structure-based drug design with equiv- ariant diffusion models. Nature Computational Science4(12), 899–909 (2024) 4

    2024

  98. [106]

    In: CVPR (2017) 5

    Simon, T., Joo, H., Matthews, I., Sheikh, Y.: Hand keypoint detection in single images using multiview bootstrapping. In: CVPR (2017) 5

    2017

  99. [107]

    arXiv preprint arXiv:2106.02733 (2021) 5

    Sosnovik, I., Moskalev, A., Smeulders, A.: Disco: accurate discrete scale convolu- tions. arXiv preprint arXiv:2106.02733 (2021) 5

    2021 arXiv

  100. [108]

    European Conference on Computer Vision (2024) 2, 4, 7, 9, 10, 11, 14, 8, 15, 16, 17, 24, 26, 27, 28, 30, 31, 32, 35

    Sun, J., Chowdhary, G.: Comusion: Towards consistent stochastic human motion prediction via motion diffusion. European Conference on Computer Vision (2024) 2, 4, 7, 9, 10, 11, 14, 8, 15, 16, 17, 24, 26, 27, 28, 30, 31, 32, 35

    2024

  101. [109]

    ACM Transactions on Graphics (ToG)24(1), 98–117 (2005) 4, 2

    Tak, S., Ko, H.S.: A physically-based motion retargeting filter. ACM Transactions on Graphics (ToG)24(1), 98–117 (2005) 4, 2

    2005

  102. [110]

    Tang, J., Yang, H., Chen, T., Hu, J.F.: Stochastic human motion prediction with memoryof action transitionand actioncharacteristic.In: Proceedingsof theCom- puter Vision and Pattern Recognition Conference. pp. 1883–1893 (2025) 5

    2025

  103. [111]

    In: European Conference on Computer Vision

    Tevet, G., Gordon, B., Hertz, A., Bermano, A.H., Cohen-Or, D.: Motionclip: Exposing human motion generation to clip space. In: European Conference on Computer Vision. pp. 358–374. Springer (2022) 5

    2022

  104. [112]

    arXiv preprint arXiv:2004.03990 (2020) 8, 9, 13, 14 EquiFusion: Kinematics-Agnostic HMP 23

    Thiede, E.H., Hy, T.S., Kondor, R.: The general theory of permutation equiv- arant neural networks and higher order graph variational encoders. arXiv preprint arXiv:2004.03990 (2020) 8, 9, 13, 14 EquiFusion: Kinematics-Agnostic HMP 23

    2004 arXiv

  105. [113]

    In: Conference on Computer Vision and Pattern Recognition (CVPR)

    Tripathi, S., Müller, L., Huang, C.H.P., Omid, T., Black, M.J., Tzionas, D.: 3D human pose estimation via intuitive physics. In: Conference on Computer Vision and Pattern Recognition (CVPR). pp. 4713–4725 (2023) 6, 11, 13

    2023

  106. [114]

    Advances in neural informa- tion processing systems30(2017) 9

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł., Polosukhin, I.: Attention is all you need. Advances in neural informa- tion processing systems30(2017) 9

    2017

  107. [115]

    arXiv preprint arXiv:1710.10903 (2017) 9

    Veličković, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., Bengio, Y.: Graph attention networks. arXiv preprint arXiv:1710.10903 (2017) 9

    2017 arXiv

  108. [116]

    In: Proceedings of the IEEE/CVF International Con- ference on Computer Vision

    Villegas, R., Ceylan, D., Hertzmann, A., Yang, J., Saito, J.: Contact-aware retar- geting of skinned motion. In: Proceedings of the IEEE/CVF International Con- ference on Computer Vision. pp. 9720–9729 (2021) 2, 4

    2021

  109. [117]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Villegas, R., Yang, J., Ceylan, D., Lee, H.: Neural kinematic networks for unsuper- vised motion retargetting. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 8639–8648 (2018) 2, 4

    2018

  110. [118]

    In: European Conference on Computer Vision

    Wad, T., Sun, Q., Pranata, S., Jayashree, K., Zhang, H.: Equivariance and invari- ance inductive bias for learning from insufficient data. In: European Conference on Computer Vision. pp. 241–258. Springer (2022) 8, 13, 14

    2022

  111. [119]

    In: Proceedings of the IEEE international conference on computer vision

    Walker, J., Marino, K., Gupta, A., Hebert, M.: The pose knows: Video forecasting by generating pose futures. In: Proceedings of the IEEE international conference on computer vision. pp. 3332–3341 (2017) 4, 5, 9, 11, 7, 8, 26, 27, 28, 30, 31, 32

    2017

  112. [120]

    Wearable technologies4, e11 (2023) 3

    Wang, H., Basu, A., Durandau, G., Sartori, M.: A wearable real-time kinetic measurement sensor setup for human locomotion. Wearable technologies4, e11 (2023) 3

    2023

  113. [121]

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

    Wang, J., Li, X., Liu, S., De Mello, S., Gallo, O., Wang, X., Kautz, J.: Zero- shot pose transfer for unrigged stylized 3d characters. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 8704– 8714 (2023) 4, 5, 2

    2023

  114. [122]

    Wang, J., Yang, F., Gou, W., Li, B., Yan, D., Zeng, A., Gao, Y., Wang, J., Zhang, R.: Freeman: Towards benchmarking 3d human pose estimation in the wild (2023) 2

    2023

  115. [123]

    Advances in Neural Information Processing Systems37, 127483–127510 (2024) 9

    Wang, Y., Hu, K., Gupta, S., Ye, Z., Wang, Y., Jegelka, S.: Understanding the role of equivariance in self-supervised learning. Advances in Neural Information Processing Systems37, 127483–127510 (2024) 9

    2024

  116. [124]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Wei, D., Sun, H., Li, B., Lu, J., Li, W., Sun, X., Hu, S.: Human joint kinematics diffusion-refinement for stochastic motion prediction. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 37, pp. 6110–6118 (2023) 4, 5, 7, 16

    2023

  117. [125]

    In: CVPR (2016) 5

    Wei, S.E., Ramakrishna, V., Kanade, T., Sheikh, Y.: Convolutional pose ma- chines. In: CVPR (2016) 5

    2016

  118. [126]

    arXiv preprint arXiv:2304.05864 (2023) 5

    Wimmer, T., Golkov, V., Dang, H.N., Zaiss, M., Maier, A., Cremers, D.: Scale- equivariant deep learning for 3d data. arXiv preprint arXiv:2304.05864 (2023) 5

    2023 arXiv

  119. [127]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Xu, C., Tan, R.T., Tan, Y., Chen, S., Wang, X., Wang, Y.: Auxiliary tasks benefit 3d skeleton-based human motion prediction. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 9509–9520 (2023) 2, 4, 5

    2023

  120. [128]

    European Conference on Computer Vision (2024) 8

    Xu, G., Tao, J., Li, W., Duan, L.: Learning semantic latent directions for accurate and controllable human motion prediction. European Conference on Computer Vision (2024) 8

    2024

  121. [129]

    arXiv preprint (2026) 6 24 C

    Xu,M.,Wang,Q.,Wen,Z.,Thien,P.D.,Li,Z.,Zhang,N.,He,X.,Zhao,W.,Gong, K., Zhang, M.: Necromancer: Breathing life into skeletons via bvh animation. arXiv preprint (2026) 6 24 C. Curreli et al

    2026

  122. [130]

    In: European Conference on Computer Vision

    Xu, S., Wang, Y.X., Gui, L.Y.: Diverse human motion prediction guided by multi- level spatial-temporal anchors. In: European Conference on Computer Vision. pp. 251–269. Springer (2022) 4

    2022

  123. [131]

    In: Proceedings of the European conference on computer vision (ECCV)

    Yan, X., Rastogi, A., Villegas, R., Sunkavalli, K., Shechtman, E., Hadap, S., Yumer, E., Lee, H.: Mt-vae: Learning motion transformations to generate multi- modal human dynamics. In: Proceedings of the European conference on computer vision (ECCV). pp. 265–281 (2018) 4

    2018

  124. [132]

    arXiv preprint arXiv:2511.02027 (2025) 3

    Yang, Z., Leite, C.S., Xiao, Y.: Strengthsense: A dataset of imu signals capturing everyday strength-demanding activities. arXiv preprint arXiv:2511.02027 (2025) 3

    2025

  125. [133]

    generation: Taming optimization dilemma in latent diffusion models

    Yao, J., Yang, B., Wang, X.: Reconstruction vs. generation: Taming optimization dilemma in latent diffusion models. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 15703–15712 (2025) 8, 35

    2025

  126. [134]

    Ying, C., Cai, T., Luo, S., Zheng, S., Ke, G., He, D., Shen, Y., Liu, T.Y.: Do transformers really perform badly for graph representation? Advances in neural information processing systems34, 28877–28888 (2021) 9, 35

    2021

  127. [135]

    IEEE Transactions on Robotics (2025) 4

    Yoon, T., Kang, D., Kim, S., Cheng, J., Ahn, M., Coros, S., Choi, S.: Spatio- temporal motion retargeting for quadruped robots. IEEE Transactions on Robotics (2025) 4

    2025

  128. [136]

    In: Computer Vision–ECCV 2020: 16th European Conference, Glas- gow, UK, August 23–28, 2020, Proceedings, Part IX 16

    Yuan, Y., Kitani, K.: Dlow: Diversifying latent flows for diverse human motion prediction. In: Computer Vision–ECCV 2020: 16th European Conference, Glas- gow, UK, August 23–28, 2020, Proceedings, Part IX 16. pp. 346–364. Springer (2020) 2, 4, 5, 7, 9, 11, 8, 15, 24, 26, 27, 28...

    2020

  129. [137]

    In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Zhang, H., Chen, Z., Xu, H., Hao, L., Wu, X., Xu, S., Zhang, Z., Wang, Y., Xiong, R.: Semantics-aware motion retargeting with vision-language models. In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2155–2164 (2024) 4, 2

    2024

  130. [138]

    IEEE transactions on pattern analysis and machine intelligence46(6), 4115–4128 (2024) 6

    Zhang, M., Cai, Z., Pan, L., Hong, F., Guo, X., Yang, L., Liu, Z.: Motiondiffuse: Text-driven human motion generation with diffusion model. IEEE transactions on pattern analysis and machine intelligence46(6), 4115–4128 (2024) 6

    2024

  131. [139]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Zhong, C., Hu, L., Zhang, Z., Ye, Y., Xia, S.: Spatio-temporal gating-adjacency gcn for human motion prediction. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 6447–6456 (2022) 7, 9

    2022

  132. [140]

    Advances in neural infor- mation processing systems36, 24966–24992 (2023) 5

    Zhou, A., Yang, K., Burns, K., Cardace, A., Jiang, Y., Sokota, S., Kolter, J.Z., Finn, C.: Permutation equivariant neural functionals. Advances in neural infor- mation processing systems36, 24966–24992 (2023) 5

    2023

  133. [141]

    arXiv preprint arXiv:2310.13039 (2023) 5

    Zhou, L., Meng, X., Liu, Z., Wu, M., Gao, Z., Wang, P.: Human pose-based estimation, tracking and action recognition with deep learning: A survey. arXiv preprint arXiv:2310.13039 (2023) 5

    2023 arXiv

  134. [142]

    fixed noise

    Zhu, W., Qiu, Q., Calderbank, R., Sapiro, G., Cheng, X.: Scaling-translation- equivariant networks with decomposed convolutional filters. Journal of machine learning research23(68), 1–45 (2022) 5 Table of Contents A Extended Discussion on Related Works and Positioning ...........

    2022

  135. [143]

    Then we trained the model with joint permutation as data augmentation

    the anisotropy was removed from the diffusion training, as it relied on fixed joint positions given by the adjacent matrix. Then we trained the model with joint permutation as data augmentation. This lead us to a baseline that employs similarly to us attention and graph convol...

    1920

  136. [144]

    1, without unified metrics

    Conventional Metrics for Tab. 1, without unified metrics. Here in Tab. 19 we see that ranking is maintained between conventional and unified metrics

  137. [145]

    2 can be found in Tab

    Full metric evaluation for MoYoga of Tab. 2 can be found in Tab. 20

  138. [146]

    5 can be found in Tab

    Full metric evaluation for the ablation on the motion parametrization as bone direction in Tab. 5 can be found in Tab. 21

  139. [147]

    data", but not seen

    Full metrics evaluation for occlusion of random limbs on AMASS in Tab. 3 can be found in Tab. 22. I Ablations and Validations on EquiFusion We validate our model through extensive experiments, investigating the training methodology and the cross-topology application. In Tab. 2...

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.