Pith. sign in

REVIEW 5 major objections 6 minor 53 references

SSD-Poser: Avatar Pose Estimation with State Space Duality from Sparse Observations

T0 review · 5 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read SSD-Poser reconstructs full-body avatar motion from only head-and-hand tracking signals, reporting state-of-the-art accuracy on AMASS with 7 ms inference and 7.34M parameters.

desk verdict A coherent Mamba-2 + attention pose estimator with useful ablations, but the AMASS split and input-synthesis protocol are underspecified enough that the SOTA accuracy claim is not yet verifiable. read the letter →

arxiv 2504.18332 v1 pith:X6QI6MKV submitted 2025-04-25 cs.CV cs.HC

Shuting Zhao , Linxin Bai , Liangjing Shao , Ye Zhang , Xinrong Chen This is my paper · ORCID
classification cs.CVcs.HC
keywords humanposeestimationsparseobservationsstatespacedualityhead-mounteddisplaytrackingfrequency-awaredecoderfull-bodymotionreconstructionAMASSreal-timeinference
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

SSD-Poser aims to prove that full-body avatar pose can be reconstructed in real time from the three sparse signals a head-mounted display already provides—head and two hands—without the heavy cost of Transformers or diffusion models. The paper's central claim is that a hybrid encoder pairing the linear-time State Space Duality (SSD) recurrence with multi-head attention, followed by a frequency-aware decoder, recovers lower-body motion accurately enough to beat prior methods on the AMASS benchmark while running at 0.007 seconds per sequence. If correct, this matters for AR/VR because it offers a lightweight backbone that could track avatars from existing HMD hardware rather than requiring IMU suits or cameras. The reported numbers include MPJRE of 2.18, MPJPE of 2.67 cm, and MPJVE of 15.25 cm/s on a 14-subset AMASS split.

What carries the argument

The load-bearing object is the hybrid State Space Attention Encoder (SSAE), which pairs a Pose State Space Block (PSSB)—a gated SSD update that propagates latent state through time—with a standard multi-head attention module, so long-range temporal dynamics are handled at linear cost while attention supplies inter-joint context. The second enabling mechanism is the Frequency-Aware Decoder (FAD), whose Frequency-Aware Feature Extractor passes features through 1x1 and 1x5 convolutions to isolate low-frequency smooth motion from high-frequency detail, then concatenates and refines them before the final joint regression. Together they convert 54-dimensional sparse HMD observations into 132-dimensional SMPL pose parameters (22 joints, 6D rotations each).

What would settle it

Record a person wearing an HMD plus a synchronized optical motion-capture suit, feed only the head and hand tracks to SSD-Poser, and compare lower-body MPJPE against the 4.80 cm reported on the AMASS 14-subset split; if the real-device lower-body error exceeds roughly twice the benchmark error, the transferability claim collapses.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that a state-space/attention hybrid can outperform dedicated generative and Transformer pose estimators on sparse-input full-body tracking. SSD-Poser encodes sequences of 96 frames of head/hand positions, rotations, and velocities through four State Space Attention Encoders (SSAE), each combining a Pose State Space Block (PSSB) based on the SSD recurrence $h_t = A_t h_{t-1} + B_t x_t$, $y_t = C_t^T h_t$ with a multi-head attention module; a Frequency-Aware Decoder then separates low- and high-frequency motion features with 1x1 and 1x5 convolutions before regressing SMPL joint rotations. The paper reports that this architecture beats the compared methods on both a 3-subset and a 14-subset AMASS protocol, with the largest gains on lower-body and position errors, while using 7.34M parameters and an average inference time of 0.007 seconds per sequence. The velocity metrics (MPJVE of 15.25 cm/s and jitter of 6.73 on the larger split) support the smoothness claim.

Load-bearing premise

The entire result depends on the premise that head and hand signals alone carry enough information to infer unconstrained lower-body motion, and that AMASS's synthetic sparse observations are a faithful stand-in for real HMD tracking noise and user behavior.

Editorial extensions

If this is right

  • Real-time avatar tracking becomes feasible on HMD hardware without external sensors: 0.007 seconds per 96-frame sequence is well below interactive frame budgets.
  • The accuracy gains concentrate in lower-body and position errors (Lower PE of 4.80 cm versus 5.37 cm for the best diffusion baseline on the 14-subset split), reducing the most visible artifact in sparse tracking.
  • The lightweight 7.34M-parameter footprint makes SSD-Poser a practical backbone for on-device or embedded AR/VR systems.
  • The frequency-aware decoder's effect on MPJVE and jitter suggests that separating low- and high-frequency motion is a useful general smoothing strategy for pose regression.

Reading between the lines

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

  • The same SSAE hybrid recipe could be applied to other sparse-sensor pose tasks, such as 4-IMU lower-body tracking, where linear-time state propagation may help long sequences more than attention alone.
  • Treating the frequency split as a learnable inductive bias suggests a testable extension: replacing the fixed 1x5 temporal kernel with a learnable or multi-scale bank should show whether the gain comes from frequency separation per se or from the extra receptive field.
  • The random 90/10 AMASS split reports synthetic noise-free HMD observations; a cleaner test would be to train on AMASS and evaluate on real HMD recordings with optical ground truth, where distribution shift is likely to lower the reported accuracy.
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

5 major / 6 minor

Summary. The paper proposes SSD-Poser, a lightweight architecture for full-body pose estimation from sparse HMD tracking signals (head and two hands). The method combines a hybrid encoder (State Space Attention Encoder, SSAE) that mixes a state-space-duality (SSD/Mamba-2) block with multi-head attention, and a Frequency-Aware Decoder (FAD) with parallel 1x1 and 1x5 convolutions to separate low- and high-frequency motion components. The model is trained end-to-end on AMASS with a loss combining rotation, position, and root orientation terms, and evaluated under two AMASS splits against prior HMD-based methods (AvatarPoser, DAP, AGRoL, KCTD, AvatarJLM, SAGE). The paper reports state-of-the-art accuracy (e.g., MPJPE 2.67 cm, MPJRE 2.18 degrees on 14 AMASS subsets) with 7.34M parameters and 0.007 s average inference per sequence, and provides ablations showing the contribution of PSSB, the SSM block, L2 loss, and the FAD module.

Significance. If the reported results are reproducible and transfer to real HMD tracking, the paper makes a useful contribution: it demonstrates that a hybrid SSD/attention encoder plus a frequency-aware decoder can achieve competitive accuracy at lower computational cost than transformer- and diffusion-based baselines, which is relevant for real-time AR/VR applications. The architecture is simple, lightweight, and the ablation study gives some support for the role of the FAD in improving smoothness (MPJVE and Jitter). However, the empirical verification is currently not strong enough to support the central SOTA claim: the evaluation protocol is underspecified in several load-bearing ways, and no uncertainty quantification is provided. The architectural novelty is incremental rather than fundamentally new, but the combination is reasonable.

major comments (5)
  1. [Section 5.3 (Evaluation Results)] The AMASS splitting protocol is not specified at the granularity needed to evaluate the results. The text says 'The datasets in both settings are randomly divided into training and test sets with 90% and 10% of the data,' but does not say whether the split is by frame, by sequence, or by subject. Because AMASS frames are temporally autocorrelated, a frame-level random split leaks near-duplicate poses into the training set and can substantially inflate accuracy; even a sequence-level split that mixes subjects can cause style leakage. Since the SOTA claim in Tables 1-2 rests entirely on this split, please specify the exact split unit, and preferably re-run the comparison with a subject/sequence-exclusive split and report results over multiple random seeds.
  2. [Section 4.1 (Problem Formulation) and Section 5.3] The synthesis of sparse inputs from AMASS is not described. Equation (6) defines the input format (positions, 6D rotations, linear and angular velocities for head and hands), but the paper never states how these signals are derived from the SMPL parameters, whether the same synthesis is applied to all baselines, or whether any sensor noise or missing-data model is used. Without this, the numbers in Tables 1-2 are not reproducible, and the claimed suitability for 'real noisy HMD tracking' is unverified. Please add a precise description of the input-generation pipeline, including coordinate frames, sampling, and any noise, and specify whether baseline methods were re-run under the identical protocol or whether numbers were taken from prior papers.
  3. [Section 5.3 and Table 2] The reported accuracy margins over the strongest baseline are small (e.g., MPJPE 2.67 vs 2.95 cm, MPJRE 2.18 vs 2.41 degrees against SAGE in Table 2), yet no error bars, multiple seeds, or significance tests are provided. Given the underspecified split and input synthesis, the claim in Section 6 that the model 'surpasses state-of-the-art approaches in reconstruction accuracy' is not currently falsifiable. Please provide confidence intervals across at least 3-5 seeds and, if possible, a per-subject breakdown to show that the improvement is consistent rather than driven by a particular split.
  4. [Section 5.4 (Ablation Study)] The hyperparameters T=96 and the number of SSAE blocks (4) appear to be selected using the same test split that is used for the final reported numbers (Tables 1-2 and the ablation Tables 5-6). If no separate validation set was held out, the reported 'optimal' configuration may be overfit to the test set. Please clarify the model-selection procedure, or re-evaluate with a validation set, and report the final test numbers only for the configuration chosen on validation.
  5. [Section 5.4 and Table 4] The text states that in the FAD ablation 'most metrics remain relatively stable,' but Table 4 shows that removing FAD degrades MPJVE from 19.32 to 27.80 and Jitter from 8.19 to 23.22, which are substantial changes. The qualitative description contradicts the quantitative result. Please correct this description and discuss the role of FAD in light of the actual numbers.
minor comments (6)
  1. [Table 3] The entry 'Avatar]LM' appears to be a typo for 'AvatarJLM;' please fix.
  2. [Section 4.3 (State Space Attention Encoder)] 'Muti-Head Attention' should be 'Multi-Head Attention.' Please also check for other spelling/grammar issues, e.g., 'in the second line, the side-kick pose (in the third line), .' in Section 5.3.
  3. [Section 5.1 (Dataset and Implementation Details)] The description of the learning rate schedule is incomplete: it says the learning rate 'decays to 3e-5 after 200000 iterations,' but does not specify the decay schedule (step, cosine, etc.). Please provide the full schedule.
  4. [Section 5.4 / Table 6] The ablation over the number of SSAE blocks reports that 5 blocks achieves the best MPJRE/MPJPE (2.37/3.11) but the paper selects 4 blocks citing 'precision, smoothness, and model efficiency.' Please justify this choice quantitatively, e.g., by reporting the inference time or parameter-vs-accuracy trade-off for each setting.
  5. [References] Several dataset references are missing full citations (e.g., ACCAD, SFU, Eyes Japan, KIT, MoVi). Please provide complete bibliographic entries consistent with the style used for other references.
  6. [Figure 4] The caption 'Visualization results of different actions compared with other state-of-the-art methods' would benefit from naming the specific actions and the color-map definition for the red error regions, which are described only in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the reported accuracy gains are held-out empirical evaluations against external AMASS ground truth, not derivations from the model's own inputs.

full rationale

SSD-Poser is an empirical model trained end-to-end on AMASS against SMPL ground truth; the paper makes no claim to derive accuracy from first principles. Section 4.5 defines the loss directly on ground-truth rotations, positions, and orientations, and Tables 1-2 compare held-out test errors against external baselines. The overlapping-author baselines KCTD [9] and DAP [11] are comparisons, not inputs to the result; no load-bearing argument cites a same-author theorem or a fitted parameter renamed as a prediction. The input definition in Eq. (6) follows AvatarPoser [22], and the SSD equations in Section 3 are external prior work. The skeptic's concerns about random-split leakage and missing noise modeling are evaluation-validity risks, not circularity. Therefore the central claim is self-contained against external data, and no circular step is exhibited.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central contribution is an architecture; the main dependencies are the SSD framework, the SMPL representation, the AMASS benchmark, and chosen hyperparameters. No new physical or mathematical entities are postulated, and no independent falsifiable handles outside the benchmark results are offered.

free parameters (6)
  • Input sequence length T = 96
    Selected via ablation in Table 5 as a trade-off between accuracy and smoothness; longer sequences reduce jitter but hurt MPJRE and MPJPE.
  • Number of SSAE blocks = 4
    Selected via ablation in Table 6; 5 blocks give lower MPJRE and MPJPE but more parameters and similar or higher jitter, so 4 is chosen as a balance.
  • Feature dimension E = 256
    No ablation is provided; chosen by hand as a standard latent capacity.
  • Loss weights alpha, beta, gamma = 1, 1, 0.02
    Reported in Section 4.5 with no ablation for gamma; these weights shape the balance between rotation, position, and root orientation terms.
  • Learning rate schedule = 3e-4 decaying to 3e-5 after 200k iterations
    Training choice reported in Section 5.1; no ablation.
  • FAFE convolution kernel sizes = 1x1 and 1x5
    Architecture choice for low- and high-frequency feature separation; no kernel-size ablation is reported.
assumptions (4)
  • domain assumption State Space Duality (SSD) framework from Mamba-2 provides efficient, content-aware sequential modeling suitable for pose estimation.
    Section 3 and Section 4.3 rely on Dao and Gu [10] for efficiency and modeling capability. This is an external prior claim the paper does not re-derive.
  • domain assumption SMPL body model with the first 22 joints and 6D rotation representation is a sufficient output representation for full-body avatars.
    Section 4.1 uses SMPL and 6D rotations; fingers are ignored. This is standard practice but is an assumption about task definition.
  • domain assumption AMASS mocap data with a random 90/10 split is a valid proxy for real HMD tracking signals.
    Section 5.1 and Section 5.3 use AMASS and synthesize sparse tracking signals as in prior work; real sensor noise and distribution shift are not modeled.
  • domain assumption Head and hand signals contain enough information to infer lower-body pose.
    This is the core task premise. The paper does not establish it beyond citing prior work that uses the same input modality.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SSD-Poser: Avatar Pose Estimation with State Space Duality from Sparse Observations." pith.science (2026). https://pith.science/paper/X6QI6MKV

@misc{pith2026250418332,
  author       = {Pith},
  title        = {Pith review of: SSD-Poser: Avatar Pose Estimation with State Space Duality from Sparse Observations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X6QI6MKV}},
  note         = {Machine review of arXiv:2504.18332}
}
read the original abstract

The growing applications of AR/VR increase the demand for real-time full-body pose estimation from Head-Mounted Displays (HMDs). Although HMDs provide joint signals from the head and hands, reconstructing a full-body pose remains challenging due to the unconstrained lower body. Recent advancements often rely on conventional neural networks and generative models to improve performance in this task, such as Transformers and diffusion models. However, these approaches struggle to strike a balance between achieving precise pose reconstruction and maintaining fast inference speed. To overcome these challenges, a lightweight and efficient model, SSD-Poser, is designed for robust full-body motion estimation from sparse observations. SSD-Poser incorporates a well-designed hybrid encoder, State Space Attention Encoders, to adapt the state space duality to complex motion poses and enable real-time realistic pose reconstruction. Moreover, a Frequency-Aware Decoder is introduced to mitigate jitter caused by variable-frequency motion signals, remarkably enhancing the motion smoothness. Comprehensive experiments on the AMASS dataset demonstrate that SSD-Poser achieves exceptional accuracy and computational efficiency, showing outstanding inference efficiency compared to state-of-the-art methods.

Figures

Figures reproduced from arXiv: 2504.18332 by the authors.

Figure 1
Figure 1. Our method takes sparse observations as inputs and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Comparison of our approach with state-of-the [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The overall framework of this work. (a) The framework of the proposed SSD-Poser model. (b) The framework of [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Visualization results of different actions compared with other state-of-the-art methods. Red regions indicate recon [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 6
Figure 6. Figure 6: These results demonstrate that the performance deteriorates [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 5
Figure 5. Figure 5: Visualization results of continuous pose sequences compared with other methods. Dashed boxes are used to highlight [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visualization results of continuous pose sequences compared with ablation methods. Dashed boxes are used to [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 29 canonical work pages

  1. [1]

    [n. d.]. ACCAD MoCap Dataset. MoCap Dataset

    work page

  2. [2]

    [n. d.]. SFU Motion Capture Database. http://mocap.cs.sfu

    work page

  3. [3]

    Carnegie-Mellon mocap database

    2007. Carnegie-Mellon mocap database. [Online]. Available: http://mocap.cs.cmu.edu/

    work page 2007

  4. [4]

    Eyes Japan MoCap Dataset

    2018. Eyes Japan MoCap Dataset. http://mocapdata.com

    work page 2018

  5. [5]

    Ijaz Akhter and Michael J Black. 2015. Pose-conditioned joint angle limits for 3D human pose reconstruction. In Proceedings of the IEEE conference on computer vision and pattern recognition . 1446–1455

    work page 2015

  6. [6]

    Sadegh Aliakbarian, Pashmina Cameron, Federica Bogo, Andrew Fitzgibbon, and Thomas J Cashman. 2022. Flag: Flow-based 3d avatar generation from sparse observations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 13253–13262

    work page 2022

  7. [7]

    Angela Castillo, Maria Escobar, Guillaume Jeanneret, Albert Pumarola, Pablo Arbeláez, Ali Thabet, and Artsiom Sanakoyeu. 2023. Bodiffusion: Diffusing sparse observations for full-body human motion synthesis. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 4221–4231

    work page 2023

  8. [8]

    Jinxiang Chai and Jessica K Hodgins. 2005. Performance animation from low- dimensional control signals. In ACM SIGGRAPH 2005 Papers. 686–696

    work page 2005

Show all 53 references
  1. [9]

    Xiaokun Dai, Xinkang Zhang, Shiman Li, and Xinrong Chen. 2024. Enhancing Motion Reconstruction From Sparse Tracking Inputs With Kinematic Constraints. IEEE Transactions on Automation Science and Engineering (2024)

    2024

  2. [10]

    Tri Dao and Albert Gu. 2024. Transformers are SSMs: Generalized models and efficient algorithms through structured state space duality. arXiv preprint arXiv:2405.21060 (2024)

    2024 arXiv

  3. [11]

    Xinhan Di, Xiaokun Dai, Xinkang Zhang, and Xinrong Chen. 2023. Dual attention poser: dual path body tracking based on attention. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 2794–2803

    2023

  4. [12]

    Andrea Dittadi, Sebastian Dziadzio, Darren Cosker, Ben Lundell, Thomas J Cash- man, and Jamie Shotton. 2021. Full-body motion from a single head-mounted device: Generating smpl poses from partial observations. In Proceedings of the IEEE/CVF International Conference on Computer...

    2021

  5. [13]

    Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, and Artsiom Sanakoyeu. 2023. Avatars grow legs: Generating smooth human motion from sparse tracking inputs with diffusion model. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog...

    2023

  6. [14]

    Han Feng, Wenchao Ma, Quankai Gao, Xianwei Zheng, Nan Xue, and Huijuan Xu. 2024. Stratified Avatar Generation from Sparse Observations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 153–163

    2024

  7. [15]

    Saeed Ghorbani, K Mahdaviani, Anne Thaler, K Kording, DJ Cook, G Blohm, and NF Troje. 2020. MoVi: A Large Multipurpose Motion and Video Dataset. arXiv

    2020

  8. [16]

    Albert Gu and Tri Dao. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752 (2023)

    2023 arXiv

  9. [17]

    Albert Gu, Karan Goel, and Christopher Ré. 2021. Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396 (2021)

    2021 arXiv

  10. [18]

    Albert Gu, Isys Johnson, Karan Goel, Khaled Saab, Tri Dao, Atri Rudra, and Christopher Ré. 2021. Combining recurrent, convolutional, and continuous-time models with linear state space layers. Advances in neural information processing systems 34 (2021), 572–585

    2021

  11. [19]

    Tairan He, Zhengyi Luo, Wenli Xiao, Chong Zhang, Kris Kitani, Changliu Liu, and Guanya Shi. 2024. Learning human-to-humanoid real-time whole-body teleoperation. arXiv preprint arXiv:2403.04436 (2024)

    2024 arXiv

  12. [20]

    Wenjun Huang and Jianguo Hu. 2024. Ml-mamba: Efficient multi-modal large language model utilizing mamba-2. arXiv preprint arXiv:2407.19832 (2024)

    2024 arXiv

  13. [21]

    Yinghao Huang, Manuel Kaufmann, Emre Aksan, Michael J Black, Otmar Hilliges, and Gerard Pons-Moll. 2018. Deep inertial poser: Learning to reconstruct human pose from sparse inertial measurements in real time. ACM Transactions on Graphics (TOG) 37, 6 (2018), 1–15

    2018

  14. [22]

    Jiaxi Jiang, Paul Streli, Huajian Qiu, Andreas Fender, Larissa Laich, Patrick Snape, and Christian Holz. 2022. Avatarposer: Articulated full-body pose tracking from sparse motion sensing. In European conference on computer vision . Springer, 443–460

    2022

  15. [23]

    Y Jiang, Y Ye, D Gopinath, J Won, AW Winkler, and CK Liu. 2022. Transformer in- ertial poser: attention-based real-time human motion reconstruction from sparse IMUs. arXiv e-prints, arXiv–2203. doi: 10.48550. arXiv preprint arXiv.2203.15720 (2022)

    2022 arXiv

  16. [24]

    Yifeng Jiang, Yuting Ye, Deepak Gopinath, Jungdam Won, Alexander W Winkler, and C Karen Liu. 2022. Transformer inertial poser: Real-time human motion reconstruction from sparse imus with simultaneous terrain generation. In SIG- GRAPH Asia 2022 Conference Papers . 1–9

    2022

  17. [25]

    Rudolph Emil Kalman. 1960. A new approach to linear filtering and prediction problems. (1960)

    1960

  18. [26]

    Diederik P Kingma. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)

    2014 arXiv

  19. [27]

    Opher Lieber, Barak Lenz, Hofit Bata, Gal Cohen, Jhonathan Osin, Itay Dalmedi- gos, Erez Safahi, Shaked Meirom, Yonatan Belinkov, Shai Shalev-Shwartz, et al

  20. [28]

    Matthew Loper, Naureen Mahmood, and Michael J Black. 2014. MoSh: motion and shape capture from sparse markers. ACM Trans. Graph. 33, 6 (2014), 220–1

    2014

  21. [29]

    Matthew Loper, Naureen Mahmood, Javier Romero, Gerard Pons-Moll, and Michael J Black. 2023. SMPL: A skinned multi-person linear model. In Seminal Graphics Papers: Pushing the Boundaries, Volume 2 . 851–866

    2023

  22. [30]

    Jun Ma, Feifei Li, and Bo Wang. 2024. U-mamba: Enhancing long-range de- pendency for biomedical image segmentation. arXiv preprint arXiv:2401.04722 (2024)

    2024 arXiv

  23. [31]

    Naureen Mahmood, Nima Ghorbani, Nikolaus F Troje, Gerard Pons-Moll, and Michael J Black. 2019. AMASS: Archive of motion capture as surface shapes. In Proceedings of the IEEE/CVF international conference on computer vision . 5442– 5451

    2019

  24. [32]

    Christian Mandery, Ömer Terlemez, Martin Do, Nikolaus Vahrenkamp, and Tamim Asfour. 2015. The KIT whole-body human motion database. In 2015 International Conference on Advanced Robotics (ICAR) . IEEE, 329–336

    2015

  25. [33]

    Harsh Mehta, Ankit Gupta, Ashok Cutkosky, and Behnam Neyshabur. 2022. Long range language modeling via gated state spaces. arXiv preprint arXiv:2206.13947 (2022)

    2022 arXiv

  26. [34]

    Meinard Müller, Tido Röder, Michael Clausen, Bernhard Eberhardt, Björn Krüger, and Andreas Weber. 2007. Mocap database hdm05. Institut für Informatik II, Universität Bonn 2, 7 (2007)

    2007

  27. [35]

    Badri N Patro and Vijay S Agneeswaran. 2024. Simba: Simplified mamba-based ar- chitecture for vision and multivariate time series. arXiv preprint arXiv:2403.15360 (2024)

    2024 arXiv

  28. [36]

    Davis Rempe, Tolga Birdal, Aaron Hertzmann, Jimei Yang, Srinath Sridhar, and Leonidas J Guibas. 2021. Humor: 3d human motion model for robust pose estimation. In Proceedings of the IEEE/CVF international conference on computer vision. 11488–11499

    2021

  29. [37]

    Noam Shazeer. 2020. Glu variants improve transformer. arXiv preprint arXiv:2002.05202 (2020)

    2020 arXiv

  30. [38]

    Leonid Sigal, Alexandru O Balan, and Michael J Black. 2010. Humaneva: Synchro- nized video and motion capture dataset and baseline algorithm for evaluation of articulated human motion. International journal of computer vision 87, 1 (2010), 4–27

    2010

  31. [39]

    Jimmy TH Smith, Andrew Warrington, and Scott W Linderman. 2022. Simplified state space layers for sequence modeling. arXiv preprint arXiv:2208.04933 (2022)

    2022 arXiv

  32. [40]

    Nikolaus F Troje. 2002. Decomposing biological motion: A framework for analysis and synthesis of human gait patterns. Journal of vision 2, 5 (2002), 2–2

    2002

  33. [41]

    Matthew Trumble, Andrew Gilbert, Charles Malleson, Adrian Hilton, and John P Collomosse. 2017. Total capture: 3D human pose estimation fusing video and inertial sensors.. In BMVC, Vol. 2. London, UK, 1–13

    2017

  34. [42]

    Aaron Van Den Oord, Oriol Vinyals, et al. 2017. Neural discrete representation learning. Advances in neural information processing systems 30 (2017)

    2017

  35. [43]

    A Vaswani. 2017. Attention is all you need. Advances in Neural Information Processing Systems (2017)

    2017

  36. [44]

    Timo Von Marcard, Bodo Rosenhahn, Michael J Black, and Gerard Pons-Moll

  37. [45]

    Alexander Winkler, Jungdam Won, and Yuting Ye. 2022. Questsim: Human motion tracking from sparse sensors with simulated avatars. In SIGGRAPH Asia 2022 Conference Papers. 1–8

    2022

  38. [46]

    Dongseok Yang, Doyeon Kim, and Sung-Hee Lee. 2021. Lobstr: Real-time lower- body pose prediction from sparse upper-body tracking signals. In Computer Graphics Forum, Vol. 40. Wiley Online Library, 265–275

    2021

  39. [47]

    Xinyu Yi, Yuxiao Zhou, Marc Habermann, Soshi Shimada, Vladislav Golyanik, Christian Theobalt, and Feng Xu. 2022. Physical inertial poser (pip): Physics- aware real-time human motion tracking from sparse inertial sensors. In Pro- ceedings of the IEEE/CVF conference on computer ...

    2022

  40. [48]

    Xinyu Yi, Yuxiao Zhou, and Feng Xu. 2021. Transpose: Real-time 3d human translation and pose estimation with six inertial sensors. ACM Transactions On Graphics (TOG) 40, 4 (2021), 1–13

    2021

  41. [49]

    Xiaozheng Zheng, Zhuo Su, Chao Wen, Zhou Xue, and Xiaojie Jin. 2023. Realistic full-body tracking from sparse observations via joint-level modeling. In Proceed- ings of the IEEE/CVF International Conference on Computer Vision . 14678–14688

    2023

  42. [50]

    Lianghui Zhu, Bencheng Liao, Qian Zhang, Xinlong Wang, Wenyu Liu, and Xinggang Wang. 2024. Vision mamba: Efficient visual representation learning with bidirectional state space model. arXiv preprint arXiv:2401.09417 (2024)

    2024 arXiv

  43. [2017]

    In Computer graphics forum, Vol

    Sparse inertial poser: Automatic 3d human pose estimation from sparse imus. In Computer graphics forum, Vol. 36. Wiley Online Library, 349–360

  44. [2020]

    arXiv preprint arXiv:2003.01888 (2020)

    2020 arXiv

  45. [2024]

    arXiv preprint arXiv:2403.19887 (2024)

    Jamba: A hybrid transformer-mamba language model. arXiv preprint arXiv:2403.19887 (2024)

    2024 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.