Pith. sign in

REVIEW 4 major objections 6 minor 38 references

PRGCN: A Graph Memory Network for Cross-Sequence Pattern Reuse in 3D Human Pose Estimation

T0 review · 4 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read Retrieving reusable pose prototypes from a graph memory bank improves monocular 3D pose estimation, yielding state-of-the-art MPJPE of 37.1 mm on Human3.6M and 13.4 mm on MPI-INF-3DHP.

desk verdict A solid incremental architecture for video-based 3D HPE with plausible SOTA numbers, but the 'cross-sequence pattern reuse' story is not actually isolated by the experiments — the memory bank is a learned dynamic filter, and the gains could be a capacity effect. read the letter →

arxiv 2510.19475 v2 pith:R6TQOVGA submitted 2025-10-22 cs.CV

Zhuoyang Xie , Yibo Zhao , Hui Huang , Riwei Wang , Zan Gao This is my paper · ORCID
classification cs.CV
keywords 3Dhumanposeestimationgraphmemorybankpatternreusestate-spacemodelMambaself-attentionMPJPEcross-domaingeneralization
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

The paper argues that existing video-based 3D pose estimators are limited because they process each sequence in isolation, ignoring that human poses repeat across sequences. PRGCN, a Pattern Reuse Graph Convolutional Network, stores a compact set of pose prototypes in a graph memory bank, retrieves them with attention, and fuses them with anatomical constraints. The authors claim this cross-sequence pattern reuse improves accuracy and cross-domain generalization, backed by new state-of-the-art results (37.1 mm MPJPE on Human3.6M, 13.4 mm on MPI-INF-3DHP). The sympathetic reader should take the claim seriously because it reframes pose estimation as retrieval and adaptation rather than per-sequence optimization.

What carries the argument

The central object is the graph memory bank M ∈ R^{K×J×J}, a learnable set of joint-connectivity prototypes. It is queried by a softmax attention over a global pooled feature descriptor, and its retrieved pattern S_new is combined with the anatomical adjacency matrix A via A' = λA + (1-λ)S_new in a memory-driven graph convolution. This mechanism is what makes cross-sequence reuse explicit and is the paper's core contribution.

What would settle it

A concrete experiment: remove the graph memory bank and replace it with a learned, input-independent matrix of the same parameter count; if MPJPE remains near 37.1 mm on Human3.6M, the claimed benefit of cross-sequence pattern reuse is not established.

Watch

Extended reading notes

Core claim

PRGCN is built on the premise that the space of physically plausible human poses is low-dimensional, so a finite set of canonical pose structures can represent it. The model learns a graph memory bank of K=48 prototypes, each a joint-connectivity matrix. Given an input sequence, an attention mechanism retrieves a weighted combination of these prototypes, which is temporally smoothed and then fused with the fixed skeletal adjacency matrix through a learnable scalar λ in a memory-driven graph convolution. The paper reports that this mechanism, combined with a dual-stream Mamba-plus-attention encoder, achieves state-of-the-art MPJPE of 37.1 mm on Human3.6M and 13.4 mm on MPI-INF-3DHP, and impro

Load-bearing premise

The load-bearing premise is that the graph memory bank stores reusable pose prototypes rather than merely being another set of learned weights; the paper does not compare against an equal-capacity network without the memory module, so the paradigm-level claim depends on that distinction.

Editorial extensions

If this is right

  • If PRGCN is right, retrieval-based pose priors can replace per-sequence optimization, yielding measurable accuracy gains on standard benchmarks.
  • The memory bank's compactness (48 prototypes for a 51-dimensional pose space) supports the manifold/covering-number explanation the paper offers.
  • The reported zero-shot cross-domain results suggest learned pose prototypes carry transferable structure, even if the gains are modest.
  • The dual-stream Mamba+attention combination provides a template for balancing local and global modeling in pose lifting.
  • Velocity loss and temporal smoothing with a gated memory state reduce jitter, as evidenced by the P-MPJPE improvements.

Reading between the lines

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

  • Editorial inference: the memory bank is static at inference, so the 'cross-sequence reuse' effect is learned only from the training split; a natural extension is to update the bank at test time, which the paper does not explore.
  • Editorial inference: because no ablation compares PRGCN to an equal-capacity network without the memory bank, part of the reported gain could come from added parameters; a parameter-matched baseline would isolate the pattern-reuse effect.
  • Editorial inference: the idea could transfer to other structured-output tasks (e.g., hand or animal pose estimation) where anatomical constraints and recurring configurations also apply.
  • Editorial inference: the paper's own conclusion admits efficiency gains of pattern reuse are not yet realized; optimizing retrieval is a clear next step.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes PRGCN, a video-based 3D human pose estimation method that combines a dual-stream Mamba/attention encoder with a graph memory bank of K pose prototypes. The memory bank is queried by an attention mechanism, the retrieved prototypes are aggregated into a dynamic J×J graph, temporally smoothed, and fused with a hard-coded anatomical adjacency matrix via a memory-driven graph convolution. The method is evaluated on Human3.6M and MPI-INF-3DHP, reporting state-of-the-art MPJPE values of 37.1 mm and 13.4 mm respectively, plus a cross-domain P-MPJPE of 131.67 mm on MPI-INF-3DHP when trained on Human3.6M. The central contribution is framed as a paradigm shift: replacing per-sequence processing with cross-sequence pattern reuse through an external memory of pose prototypes.

Significance. If the reported numbers are reproducible, PRGCN would be a competitive 3D pose lifter, and the memory-driven graph convolution is an interesting architectural idea. The paper has strengths: it evaluates on multiple benchmarks, provides qualitative analyses, and its component ablations show monotonic gains when modules are added. However, the load-bearing claim that cross-sequence pattern reuse is the mechanism behind the gains is not isolated experimentally: the memory bank is a learned dynamic graph filter whose parameters are trained by the same global loss as any other layer, and no equal-capacity baseline is provided. The main results also lack statistical significance information, and Table 5 contradicts its own prose. These issues make the paradigm-level claim conditional rather than established.

major comments (4)
  1. [§3.2, Eq. (8), Algorithm 1; Table 4] The memory bank M ∈ R^{K×J×J} is updated only by the final pose loss via Eq. (2) and is frozen at inference; the retrieval operation in Eq. (8) is a softmax-weighted sum of learned matrices. This is functionally a dynamic graph filter or mixture-of-experts, not a demonstrated external memory that reuses patterns across sequences. Table 4 adds 'Pattern Reuse' on top of Step 2, but this also adds parameters; no control matches the capacity and computation of the memory module without the 'reuse' interpretation. Consequently, the paper's central claim that cross-sequence pattern reuse produces the SOTA results is unsupported. I request an equal-capacity baseline (same parameter count/compute) replacing the memory bank with a non-memory dynamic filter, and an ablation that isolates the number of learned parameters from the memory mechanism.
  2. [§4.4.2, Table 5] The prose states that increasing K from 48 to 64 'only slightly improves to 37.5mm,' but Table 5 lists K=64 at 37.5mm, which is worse than K=48's 37.1mm. The same paragraph claims steady improvement from K=16 to K=48 while ignoring the interleaved compression-ratio rows (ratio 2 gives 38.0mm, ratio 6 gives 37.2mm). These inconsistencies undermine the claimed saturation at K=48 and the 'covering number' justification. The numbers should be corrected and a clean K sweep with all other hyperparameters held fixed should be reported.
  3. [Tables 1–3] The state-of-the-art and cross-domain claims rest on differences of 0.8mm on Human3.6M (Table 1), 0.4mm under Protocol 1†, and 0.9–1.8% in the cross-domain comparison (Table 3). No standard deviations, multiple seeds, or significance tests are reported anywhere. In Table 3, the per-sequence standard deviations are about 40–60mm, so the overall 131.67mm vs. 134.02mm difference is plausibly within run-to-run noise. Please provide mean±std over at least three seeds for the main results and distinguish statistically supported improvements from trend-level ones.
  4. [Algorithm 1, Eq. (9)] The temporal smoothing/gating operation with S_prev is the only part of the algorithm that could introduce cross-chunk or cross-sequence state, but it is never ablated or analyzed. Table 4 only ablates the full 'Pattern Reuse' module. Since the manuscript claims that temporal smoothing suppresses jitter and is part of the pattern-reuse mechanism, I request ablations with and without the gating/update step and with and without S_prev to determine whether this component contributes to the reported gains.
minor comments (6)
  1. [Table 1] The PRGCN row is labeled 'PRGCN-' in the main table; this is likely a typo and should be corrected.
  2. [Eq. (16)] The notation α = Softmax(W_gate[...]) produces a two-element vector but the text then uses α_0 and α_1 without explicitly defining the indexing; please clarify.
  3. [§3.2] The initialization of the memory bank is justified by 'competitive learning theory,' but the actual update is plain backpropagation via Eq. (2). This theoretical grounding is not used anywhere; either connect it to the training procedure or remove the claim.
  4. [Figure 6] The claim of a 'very low mean pairwise correlation' among selected prototypes is asserted without numbers. If this is an important evidence for non-redundancy, provide the measured value.
  5. [§4.4.2] The statement that the required number of prototypes grows as O(d log n) by 'Covering Number theory' is presented as a mathematical justification but no derivation or citation for this specific bound is given. Please label it as a heuristic or provide a formal reference.
  6. [Availability] No code release or reproducibility statement is mentioned. Given the SOTA claims, a code/data statement would strengthen the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: benchmark numbers are measured on held-out data, not reconstructed from fitted inputs.

full rationale

The paper's headline claims are empirical: 37.1mm MPJPE on Human3.6M and 13.4mm on MPI-INF-3DHP are evaluated under standard held-out protocols (S1,S5-S8 train; S9,S11 test for Human3.6M, per Section 4.1). The graph memory bank M is trained end-to-end with the same MPJPE/velocity loss used for the network (Eq. 2 and Eqs. 20-22); nothing in the test set is used to define or update the prototypes. Eq. 8 (M_r = sum_k w_k M_k) and Eq. 10 (A' = lambda A + (1-lambda) S_new) implement a learned, input-dependent dynamic graph filter; labeling this 'cross-sequence pattern reuse' is an interpretation of the mechanism, not a derivation that reduces to its own outputs. The ablations in Table 4 measure incremental MPJPE changes on the held-out test split, so the component gains are not fitted to reproduce the final number. The covering-number argument for K=48 is offered after empirical selection and is not used to predict the benchmark results. The paper itself notes limitations ('relying solely on the memory mechanism is not sufficient to completely solve the domain shift problem', Section 4.3.3; efficiency gains 'not yet fully realized', Conclusion), and these limitations concern experimental support rather than circularity. The lack of a matched-capacity ablation weakens the causal attribution of gains to 'pattern reuse' specifically, but that is a confound/interpretation issue, not a circular step. No load-bearing self-citation or imported uniqueness/ansatz is present. Under the stated hard rules, no circular step can be identified.

Assumptions & free parameters 5 free parameters · 5 assumptions · 1 invented entities

The central empirical claim rests mostly on standard supervised-learning assumptions: datasets are representative, a low-dimensional pose manifold exists, and the architectural ablations isolate the contribution of each component. The paper adds several hand-tuned hyperparameters (K, compression ratio, loss weights) and assumes without direct measurement that the pose manifold is coverable by about 48 prototypes. No new physical or mathematical entity is introduced; the graph memory bank is an internal mechanism.

free parameters (5)
  • number of memory prototypes K = 48
    Chosen by ablation on Human3.6M (Table 5). The prose invokes covering-number theory to justify a compact K, but the actual value is empirical and K=64 is worse.
  • temporal compression ratio = 3
    Chosen by ablation (Table 5): ratio 2 gives 38.0 mm, ratio 6 gives 37.2 mm, ratio 3 gives 37.1 mm.
  • velocity loss weight lambda_v = not reported
    Appears in Eq. 20, but its value is never given, so the training objective cannot be exactly reproduced.
  • memory/anatomy fusion scalar lambda = learned; not reported
    Single learnable scalar in Eq. 10 balancing static skeleton adjacency A and retrieved memory S_new; final value is not reported.
  • adaptive fusion weights alpha = learned; not reported
    Used in Eqs. 4 and 16 to fuse spatial/temporal and Mamba/attention streams; learned values are not reported.
assumptions (5)
  • standard math The covering-number argument O(d log n) applies to the human pose manifold and justifies a small prototype set.
    Cited in Section 4.4.2 as a theoretical basis for K=48, but the paper never estimates the manifold dimension d or a covering bound, so the quantitative conclusion is heuristic.
  • domain assumption The space of anatomically plausible human poses is a low-dimensional manifold coverable by about 48 prototypes.
    Invoked in Sections 1, 3.1, and 4.4.2 via reference [34]; not directly measured, and the prototype count is selected by ablation.
  • domain assumption Common motion patterns recur across subjects and sequences, so a small fixed bank is sufficient.
    Core premise of 'cross-sequence pattern reuse' stated in Sections 2.3 and 3.2; no prototype-retrieval frequency or coverage analysis is provided.
  • domain assumption A fixed skeletal adjacency matrix A plus one learnable scalar lambda is sufficient to enforce anatomical plausibility.
    Used in Eq. 10 and 11; assumes the hard-coded topology captures the geometric constraints needed to prevent implausible poses.
  • domain assumption Standard benchmark protocols and Stacked Hourglass 2D detections are accepted as faithful evaluation inputs.
    Evaluation in Section 4.1 follows [14] and prior work; the paper does not independently verify detector noise or protocol details.
invented entities (1)
  • Graph memory bank of pose prototypes M_k in R^{J x J}
    purpose: Stores and retrieves learned joint-connectivity patterns to condition 3D pose regression (Eqs. 7-10).
    An internal learned representation with no falsifiable handle outside the paper's benchmark results. It is not shown to accumulate knowledge at test time and is only trained via the final prediction loss.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PRGCN: A Graph Memory Network for Cross-Sequence Pattern Reuse in 3D Human Pose Estimation." pith.science (2026). https://pith.science/paper/R6TQOVGA

@misc{pith2026251019475,
  author       = {Pith},
  title        = {Pith review of: PRGCN: A Graph Memory Network for Cross-Sequence Pattern Reuse in 3D Human Pose Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R6TQOVGA}},
  note         = {Machine review of arXiv:2510.19475}
}
read the original abstract

Monocular 3D human pose estimation remains a fundamentally ill-posed inverse problem due to the inherent depth ambiguity in 2D-to-3D lifting. While contemporary video-based methods leverage temporal context to enhance spatial reasoning, they operate under a critical paradigm limitation: processing each sequence in isolation, thereby failing to exploit the strong structural regularities and repetitive motion patterns that pervade human movement across sequences. This work introduces the Pattern Reuse Graph Convolutional Network (PRGCN), a novel framework that formalizes pose estimation as a problem of pattern retrieval and adaptation. At its core, PRGCN features a graph memory bank that learns and stores a compact set of pose prototypes, encoded as relational graphs, which are dynamically retrieved via an attention mechanism to provide structured priors. These priors are adaptively fused with hard-coded anatomical constraints through a memory-driven graph convolution, ensuring geometrical plausibility. To underpin this retrieval process with robust spatiotemporal features, we design a dual-stream hybrid architecture that synergistically combines the linear-complexity, local temporal modeling of Mamba-based state-space models with the global relational capacity of self-attention. Extensive evaluations on Human3.6M and MPI-INF-3DHP benchmarks demonstrate that PRGCN establishes a new state-of-the-art, achieving an MPJPE of 37.1mm and 13.4mm, respectively, while exhibiting enhanced cross-domain generalization capability. Our work posits that the long-overlooked mechanism of cross-sequence pattern reuse is pivotal to advancing the field, shifting the paradigm from per-sequence optimization towards cumulative knowledge learning.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 3 canonical work pages

  1. [1]

    In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (2018)

    Luvizon, D.C., Picard, D., Tabia, H.: 2d/3d pose estimation and action recogni- tion using multitask deep learning. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (2018). https://doi.org/10.1109/cvpr.2018.00539

    arXiv 2018

  2. [2]

    Journal of Visual Communication and Image Representation76, 103055 (2021) https://doi.org/10.1016/j.jvcir.2021.103055

    Song, L., Yu, G., Yuan, J., Liu, Z.: Human pose estimation and its application to action recognition: A survey. Journal of Visual Communication and Image Representation76, 103055 (2021) https://doi.org/10.1016/j.jvcir.2021.103055

    arXiv 2021

  3. [3]

    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 (2023)

    2023

  4. [4]

    IEEE Transactions on Industrial Informatics 18(10), 7107–7117 (2022) https://doi.org/10.1109/tii.2022.3143605

    Liu, H., Liu, T., Zhang, Z., Sangaiah, A.K., Yang, B., Li, Y.: Arhpe: Asym- metric relation-aware representation learning for head pose estimation in indus- trial human–computer interaction. IEEE Transactions on Industrial Informatics 18(10), 7107–7117 (2022) https://doi.org/10.1109/tii.2022.3143605

    arXiv 2022

  5. [5]

    Neurocomputing596, 128049 (2024) https://doi.org/ 10.1016/j.neucom.2024.128049

    Liu, Y., Qiu, C., Zhang, Z.: Deep learning for 3d human pose estimation and mesh recovery: A survey. Neurocomputing596, 128049 (2024) https://doi.org/ 10.1016/j.neucom.2024.128049

    arXiv 2024

  6. [6]

    IEEE Transactions on Visualization and Computer Graphics 22(12), 2633–2651 (2016) https://doi.org/10.1109/tvcg.2015.2513408

    Marchand, E., Uchiyama, H., Spindler, F.: Pose estimation for augmented reality: A hands-on survey. IEEE Transactions on Visualization and Computer Graphics 22(12), 2633–2651 (2016) https://doi.org/10.1109/tvcg.2015.2513408

    arXiv 2016

  7. [7]

    BMC Geriatrics 24(1) (2024) https://doi.org/10.1186/s12877-024-05188-7

    He, S., Meng, D., Wei, M., Guo, H., Yang, G., Wang, Z.: Proposal and validation of a new approach in tele-rehabilitation with 3d human posture estimation: a randomized controlled trial in older individuals with sarcopenia. BMC Geriatrics 24(1) (2024) https://doi.org/10.1186/s12877-024-05188-7

    work page doi:10.1186/s12877-024-05188-7 2024

  8. [8]

    Frontiers in Computer Science5 (2023) https://doi.org/10.3389/fcomp.2023.1153160

    Avogaro, A., Cunico, F., Rosenhahn, B., Setti, F.: Markerless human pose esti- mation for biomedical applications: a survey. Frontiers in Computer Science5 (2023) https://doi.org/10.3389/fcomp.2023.1153160

    arXiv 2023

Show all 38 references
  1. [9]

    5935–5946 (2025)

    Yeung, C., Suzuki, T., Tanaka, R., Yin, Z., Fujii, K.: Athletepose3d: A benchmark dataset for 3d human pose estimation and kinematic validation in athletic move- ments, pp. 5935–5946 (2025). https://doi.org/10.1109/cvprw67362.2025.00592

    2025

  2. [10]

    Sports Engineering27(1) (2024) https://doi.org/10.1007/s12283-024-00460-w

    Fukushima, T., Blauberger, P., Guedes Russomanno, T., Lames, M.: The poten- tial of human pose estimation for motion capture in sports: a validation study. Sports Engineering27(1) (2024) https://doi.org/10.1007/s12283-024-00460-w

    2024 doi

  3. [11]

    In: European Conference on Computer Vision, pp

    Newell, A., Yang, K., Deng, J.: Stacked hourglass networks for human pose estimation. In: European Conference on Computer Vision, pp. 483–499 (2016) 24

    2016

  4. [12]

    In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Sun, K., Xiao, B., Liu, D., Wang, J.: Deep high-resolution representation learn- ing for human pose estimation. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5686–5696 (2019). https://doi.org/ 10.1109/cvpr.2019.00584

    2019

  5. [13]

    In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Chen, Y., Wang, Z., Peng, Y., Zhang, Z., Yu, G., Sun, J.: Cascaded pyramid network for multi-person pose estimation. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 7103–7112 (2018)

    2018

  6. [14]

    In: 2017 IEEE International Conference on Com- puter Vision (ICCV), pp

    Martinez, J., Hossain, R., Romero, J., Little, J.J.: A simple yet effective baseline for 3d human pose estimation. In: 2017 IEEE International Conference on Com- puter Vision (ICCV), pp. 2659–2668 (2017). https://doi.org/10.1109/iccv.2017. 288

    2017 doi

  7. [15]

    In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Pavllo, D., Feichtenhofer, C., Grangier, D., Auli, M.: 3d human pose estima- tion in video with temporal convolutions and semi-supervised training. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 7745–7754 (2019). https://doi.org/10.1109/cvp...

    2019

  8. [16]

    In: 2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillance (A VSS), pp

    Bouazizi, A., Kressel, U., Belagiannis, V.: Learning temporal 3d human pose estimation with pseudo-labels. In: 2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillance (A VSS), pp. 1–8 (2021). https: //doi.org/10.1109/avss52988.2021.9663755

    2021

  9. [17]

    In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pp

    Cai, Y., Ge, L., Liu, J., Cai, J., Cham, T.-J., Yuan, J., Thalmann, N.M.: Exploit- ing spatial-temporal relationships for 3d pose estimation via graph convolutional networks. In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 2272–2281 (2019). https://do...

    2019

  10. [18]

    In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Zhao, L., Peng, X., Tian, Y., Kapadia, M., Metaxas, D.N.: Semantic graph convo- lutional networks for 3d human pose regression. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3420–3430 (2019). https://doi.org/10.1109/cvpr.2019.00354

    2019

  11. [19]

    In: 2023 IEEE/CVF International Conference on Com- puter Vision (ICCV), pp

    Yu, B.X.B., Zhang, Z., Liu, Y., Zhong, S.-H., Liu, Y., Chen, C.W.: Gla-gcn: Global-local adaptive graph convolutional network for 3d human pose estimation from monocular video. In: 2023 IEEE/CVF International Conference on Com- puter Vision (ICCV), pp. 8784–8795 (2023). https:...

    2023

  12. [20]

    In: Advances in Neural Information Processing Systems, vol

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, L., Polosukhin, I.: Attention is all you need. In: Advances in Neural Information Processing Systems, vol. 30 (2017)

    2017

  13. [22]

    In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Zhang, J., Tu, Z., Yang, J., Chen, Y., Yuan, J.: Mixste: Seq2seq mixed spatio- temporal encoder for 3d human pose estimation in video. In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 13222– 13232 (2022). https://doi.org/10.1109/cvpr52688.2022.01288

    2022

  14. [23]

    In: 2023 IEEE/CVF International Conference on Computer Vision (ICCV), pp

    Zhu, W., Ma, X., Liu, Z., Liu, L., Wu, W., Wang, Y.: Motionbert: A uni- fied perspective on learning human motion representations. In: 2023 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 15039–15053 (2023). https://doi.org/10.1109/iccv51070.2023.01385

    2023

  15. [25]

    In: 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Zhao, Q., Zheng, C., Liu, M., Wang, P., Chen, C.: Poseformerv2: Exploring frequency domain for efficient and robust 3d human pose estimation. In: 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8877–8886 (2023). https://doi.org/10.1109/cvpr52729...

    2023

  16. [26]

    IEEE Transactions on Circuits and Systems for Video Technology32(1), 198–209 (2022) https://doi

    Chen, T., Fang, C., Shen, X., Zhu, Y., Chen, Z., Luo, J.: Anatomy-aware 3d human pose estimation with bone-based pose decomposition. IEEE Transactions on Circuits and Systems for Video Technology32(1), 198–209 (2022) https://doi. org/10.1109/tcsvt.2021.3057267

    2022

  17. [28]

    Proceedings of the AAAI Conference on Artificial Intelligence39(5), 5478–5486 (2025) https://doi.org/10

    Liu, J., Liu, M., Liu, H., Li, W.: Tcpformer: Learning temporal correlation with implicit pose proxy for 3d human pose estimation. Proceedings of the AAAI Conference on Artificial Intelligence39(5), 5478–5486 (2025) https://doi.org/10. 1609/aaai.v39i5.32583

    2025

  18. [30]

    Gu, A., Goel, K., R´ e, C.: Efficiently modeling long sequences with structured state spaces (2021) https://doi.org/10.48550/ARXIV.2111.00396

    2021 doi arXiv

  19. [31]

    Gu, A., Dao, T.: Mamba: Linear-time sequence modeling with selective state spaces (2023) https://doi.org/10.48550/ARXIV.2312.00752 26

    2023 doi arXiv

  20. [32]

    In: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp

    Mondal, A., Alletto, S., Tome, D.: Hummuss: Human motion understanding using state space models. In: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2318–2330 (2024). https://doi.org/10.1109/ cvpr52733.2024.00225

    2024

  21. [33]

    Proceedings of the AAAI Conference on Artificial Intelligence39(10), 10248–10256 (2025) https://doi.org/10.1609/aaai.v39i10.33112

    Zhang, X., Bao, Q., Cui, Q., Yang, W., Liao, Q.: Pose magic: Efficient and temporally consistent human pose estimation with a hybrid mamba-gcn net- work. Proceedings of the AAAI Conference on Artificial Intelligence39(10), 10248–10256 (2025) https://doi.org/10.1609/aaai.v39i10.33112

    2025 doi

  22. [34]

    Nature Neuroscience5(11), 1226–1235 (2002) https://doi.org/10.1038/ nn963

    Todorov, E., Jordan, M.I.: Optimal feedback control as a theory of motor coor- dination. Nature Neuroscience5(11), 1226–1235 (2002) https://doi.org/10.1038/ nn963

    2002

  23. [35]

    In: 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pp

    Zheng, C., Zhu, S., Mendieta, M., Yang, T., Chen, C., Ding, Z.: 3d human pose estimation with spatial and temporal transformers. In: 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 11636–11645 (2021). https://doi.org/10.1109/iccv48922.2021.01145

    2021

  24. [36]

    In: 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (W ACV) (2024)

    Mehraban, S., Adeli, V., Taati, B.: Motionagformer: Enhancing 3d human pose estimation with a transformer-gcnformer network. In: 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (W ACV) (2024). https://doi. org/10.1109/wacv57701.2024.00677

    2024

  25. [37]

    48550/ARXIV.1410.3916 [cs.AI]

    Weston, J., Chopra, S., Bordes, A.: Memory networks (2015) https://doi.org/10. 48550/ARXIV.1410.3916 [cs.AI]

    2015 doi arXiv

  26. [38]

    Cognitive Science9(1), 75–112 (1985) https://doi.org/10.1016/s0364-0213(85) 80010-0

    RUMELHART, D., ZIPSER, D.: Feature discovery by competitive learning. Cognitive Science9(1), 75–112 (1985) https://doi.org/10.1016/s0364-0213(85) 80010-0

    1985 doi

  27. [39]

    IEEE Transactions on Pattern Analysis and Machine Intelligence36(7), 1325– 1339 (2014) https://doi.org/10.1109/tpami.2013.248

    Ionescu, C., Papava, D., Olaru, V., Sminchisescu, C.: Human3.6m: Large scale datasets and predictive methods for 3d human sensing in natural environments. IEEE Transactions on Pattern Analysis and Machine Intelligence36(7), 1325– 1339 (2014) https://doi.org/10.1109/tpami.2013.248

    2014 doi

  28. [40]

    In: 2017 International Conference on 3D Vision (3DV) (2017)

    Mehta, D., Rhodin, H., Casas, D., Fua, P., Sotnychenko, O., Xu, W., Theobalt, C.: Monocular 3d human pose estimation in the wild using improved cnn supervision. In: 2017 International Conference on 3D Vision (3DV) (2017). https://doi.org/ 10.1109/3dv.2017.00064

    2017

  29. [41]

    614–631 (2018)

    Marcard, T., Henschel, R., Black, M.J., Rosenhahn, B., Pons-Moll, G.: Recovering Accurate 3D Human Pose in the Wild Using IMUs and a Moving Camera, pp. 614–631 (2018). https://doi.org/10.1007/978-3-030-01249-6 37

    2018 doi

  30. [42]

    degree in com- puter science at Wenzhou University

    Cui, H., Hayama, T.: Hgmamba: Enhancing 3d human pose estimation with a hypergcn-mamba network (2025) https://doi.org/10.48550/ARXIV.2504.06638 27 Author Biographies Zhuoyang Xieis currently pursuing his M.S. degree in com- puter science at Wenzhou University. He received the ...

    2025 doi arXiv

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.