Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

BioPose: Biomechanically-accurate 3D Pose Estimation from Monocular Videos

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Monocular video can match multi-camera biomechanical pose accuracy.

desk verdict A promising biomechanics pipeline whose headline numbers don't yet support the integrated system, because NeurIK was trained on HMR2.0 markers while the paper reports MQ-HMR+NeurIK. read the letter →

arxiv 2501.07800 v1 pith:X4IDAJDP submitted 2025-01-14 cs.CV cs.AIcs.LG

Farnoosh Koleini , Muhammad Usama Saleem , Pu Wang , Hongfei Xue , Ahmed Helmy , Abbey Fenwick This is my paper · ORCID
classification cs.CVcs.AIcs.LG
keywords biomechanicalposeestimationmonocular3DhumanneuralinversekinematicsvirtualmarkersSMPLOpenSimmulti-querytransformer2Drefinement
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

This paper tries to establish that biomechanically accurate 3D human pose can be estimated from ordinary monocular video, matching or beating lab-grade multi-camera marker-based systems. To do this, it chains a multi-query deformable transformer that recovers a 3D body mesh (MQ-HMR) with a neural inverse kinematics network (NeurIK) that treats the mesh vertices as virtual markers and regresses joint angles and bone scales under an OpenSim biomechanical skeleton. A 2D-informed refinement step further aligns the predicted 3D structure with 2D keypoints at inference time. The reported joint-angle mean absolute error is 2.84 degrees on BML-MoVi and 3.19 degrees on OpenCap, compared with 4.50 degrees for the multi-camera OpenCap system. If correct, this would bring biomechanical analysis to clinics, sports, and home settings without markers or capture labs.

What carries the argument

The load-bearing mechanism is the virtual-marker bridge: a fixed set of SMPL mesh vertex positions, reported as $X^{exp}_{VM}\in\mathbb{R}^{142\times 3}$, is fed into NeurIK as though they were physical skin markers. NeurIK combines a spatial 1-D convolution encoder with a temporal transformer encoder, and is trained with losses on joint positions, marker positions, body scale, and joint angles, all filtered through a forward-kinematics layer built on an OpenSim full-body skeletal model that constrains degrees of freedom. The other key component is the MQ-HMR decoder, a multi-query deformable cross-attention mechanism over multi-scale vision-transformer features, whose pose-query tokens are also optimized at inference time to minimize the reprojection error against 2D keypoints. Together, these parts convert image appearance into anatomically constrained joint angles without an explicit marker set.

What would settle it

Run BioPose on a motion-capture dataset with simultaneous marker-based ground truth, then replace its virtual markers with an anatomically defined marker set of the same size (or perturb the chosen vertex indices): if the joint-angle error changes substantially, the mapping is tied to the arbitrary vertices rather than to genuine biomechanics. Equivalently, if the paper's undisclosed vertex indices cannot be used to reproduce the reported 2.84-degree BML-MoVi and 3.19-degree OpenCap errors, the central claim fails reproducibility.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a monocular pipeline can produce biomechanically accurate kinematics: the MQ-HMR model recovers an SMPL mesh, 142 of its vertices are treated as virtual markers, and the NeurIK spatio-temporal network maps those markers to the joint-angle and bone-scale parameters of an OpenSim full-body skeletal model. The forward-kinematics layer in NeurIK enforces anatomical degrees of freedom and range-of-motion limits, so the predicted poses are constrained to be physically plausible. The authors assert that this system outperforms prior single-camera biomechanical methods and achieves joint-angle errors below those of the multi-camera OpenCap markerless system on the OpenCap dataset. The central claim is that the virtual-marker representation, together with learned inverse kinematics, is sufficient to replace marker-based motion capture for biomechanical pose accuracy.

Load-bearing premise

The fixed set of 142 SMPL mesh vertices used as virtual markers must sit at anatomically stable sites that behave like real skin markers when fed to OpenSim inverse kinematics, but the paper neither specifies which vertices these are nor validates that equivalence.

Editorial extensions

If this is right

  • Biomechanical analysis for physical therapy, sports science, and ergonomics could be performed from single-camera video rather than marker-based labs, which is the motivation the paper states.
  • Because the virtual markers come from the SMPL mesh, any improvement in mesh recovery, such as a better human-mesh-recovery backbone, should directly translate into better biomechanical pose accuracy.
  • The 2D-informed refinement's gains appear after only a few iterations (5 to 10), so the accuracy improvement can be traded against inference latency.
  • Cross-dataset results on OpenCap and BEDLAM, with no training on those datasets, indicate that the learned marker-to-angle mapping is not tied to a single motion-capture studio.

Reading between the lines

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

  • The paper never specifies which 142 SMPL vertices act as virtual markers; if they are not anatomically stable sites, the learned mapping could be an arbitrary regression rather than a true biomechanical inverse kinematics, so the cross-dataset numbers would not transfer to clinical use.
  • The 4.50-degree OpenCap multi-camera baseline may not be directly comparable because OpenCap's error is reported on its own pipeline and test distribution; a head-to-head evaluation on identical motions and subjects would make the claimed superiority concrete.
  • A diagnostic experiment the authors did not run is to feed ground-truth SMPL meshes, or meshes fitted to marker-based motion capture, into NeurIK; the residual error would isolate how much of the 2.84-degree total comes from mesh recovery versus from the marker-to-angle mapping.
  • The 2D-informed refinement relies on OpenPose detections, so its benefit will likely shrink in videos with strong occlusion or motion blur, and the reported average inference time of about 0.6 seconds per image at 10 iterations may limit real-time applications.
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 BioPose, a three-stage pipeline for biomechanically accurate 3D pose estimation from monocular video. MQ-HMR recovers SMPL meshes using a ViT encoder and a multi-query deformable transformer; NeurIK converts 142 mesh vertices (treated as 'virtual markers') into OpenSim-style joint angles and body scales using a spatial-convolution/temporal-transformer network; and a 2D-informed refinement stage optimizes MQ-HMR's pose query tokens at inference to better match 2D keypoints. Experiments report MQ-HMR improvements over HMR2.0 and TokenHMR on Human3.6M, 3DPW, and EMDB, and claim NeurIK-based angle errors of 2.84 degrees on BML-MoVi and 3.19 degrees on OpenCap, which are presented as comparable to or better than the multi-camera OpenCap baseline of 4.50 degrees.

Significance. If the reported results hold, BioPose would be a meaningful step toward making biomechanical joint-angle estimation accessible from monocular video, with clear applications in rehabilitation, sports science, and ergonomics. The decomposition into mesh recovery, neural inverse kinematics, and inference-time 2D refinement is sensible, and the paper includes extensive ablations over pose-token counts, feature resolutions, refinement iterations, temporal window sizes, and loss terms. On the positive side, there is no equation-level circularity: NeurIK is a supervised regressor from virtual markers to OpenSim-derived angles, and the 2D refinement is a test-time fitting procedure. The main weaknesses are reproducibility and verification gaps: no code is provided, no error bars are reported, and the appendix reveals that NeurIK was trained on HMR2.0-generated virtual markers rather than MQ-HMR outputs, which directly affects the attribution of the headline results. The benchmarks also measure agreement with OpenSim IK outputs, so the numbers reflect consistency with a particular biomechanical processing pipeline rather than ground-truth skeletal motion.

major comments (4)
  1. [Appendix §1.2, §3.3, Table 2] Appendix §1.2 states that to generate NeurIK's training input, 'the SMPL mesh for each video frame was recovered using a test-time optimized HMR 2.0 model,' yet the system evaluated in Table 2 is 'MQ-HMR + NeurIK' and 'MQ-HMR + NeurIK†', whose virtual markers come from MQ-HMR meshes. Because NeurIK is a supervised regressor from marker positions to joint angles, this train/test distribution shift at the component interface means the BML-MoVi, BEDLAM, and OpenCap results in Table 2 cannot be attributed to the proposed MQ-HMR+NeurIK pipeline as written. The authors must either retrain NeurIK on MQ-HMR outputs or report the system as HMR2.0+NeurIK and temper the claims accordingly.
  2. [Table 2] Table 2 contains two 'HMR2.0 + NeurIK (Ours)' rows with different values (e.g., OpenCap MAE angle 3.41 versus 3.21, and BML-MoVi MAE body 5.86 versus 4.11). The text in §4.3.2 computes the 6.4% OpenCap improvement from the 3.41 baseline, but the other row's 3.21 is only 0.6% worse than the reported 3.19 for MQ-HMR+NeurIK†. This duplication makes the headline comparison ambiguous; the authors should clarify which row is the correct baseline and report improvements against the best available baseline.
  3. [Table 5] The feature-resolution ablation in Table 5 has duplicate and inconsistent rows: the '✓ ✓ ✓ ✓' condition appears with EMDB MPJPE 91.5 in the first row but 92.5 in the fourth row, and the '✓ ✓ ✓ ✗' condition appears twice with different values (72.5/92.9 and 70.1/90.5). Because the text uses this table to support the claimed multi-scale gains (3.5 mm and 2.9 mm reductions), these inconsistencies undermine a central design choice; the table must be corrected and the ablations re-reported consistently.
  4. [Appendix §1.2, §3.3] The virtual marker set is never specified or validated: §3.3 uses M=142 virtual markers, but the appendix only says that 'vertex indices of virtual markers on the SMPL mesh were used to calculate marker locations.' No list of SMPL vertex indices is given, nor any evidence that these locations correspond to anatomically stable bony-landmark sites under the OpenSim model. Without this, the claim that NeurIK's learned marker-to-angle mapping is biomechanically meaningful, and the cross-dataset numbers on OpenCap and BEDLAM, cannot be interpreted.
minor comments (6)
  1. [Table 2 caption] The caption of Table 2 says underlined values highlight the fourth-best performance, but the table as rendered has no underlined entries; clarify which baseline is meant.
  2. [Table 2] The OpenCap Multi-Camera row reports only MAE angle (4.50), so the comparison is partial; the other metrics should either be filled in or their absence justified.
  3. [Appendix tables] The appendix tables are numbered Table 1 and Table 2, which conflicts with the main-text Table 1 and Table 2; the appendix tables should be renumbered.
  4. [§1.11 and throughout] The text refers to 'Neura lIK' in §1.11 and uses 'BMLmovi' inconsistently alongside 'BML-MoVi'; standardize the terminology.
  5. [General] No code or checkpoints are provided, and no error bars or multiple-seed statistics are reported; given the duplicated rows in the tables, a code release or a public evaluation protocol would greatly increase confidence in the results.
  6. [References] Reference [37] is cited for OpenSim but appears to be a survey or e-health guide rather than the original OpenSim software paper; the authors should cite the primary OpenSim publication.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the pipeline stages are independently supervised regressors and the OpenSim-based target is a modeling choice, not a by-construction identity.

full rationale

BioPose's derivation chain is not circular. MQ-HMR is trained with standard SMPL, 3D-keypoint, and 2D-keypoint losses against external annotations; NeurIK is a supervised regressor from SMPL mesh vertices (virtual markers) to OpenSim/BSK joint angles and scales, trained with losses Lj, Lm, Ls, and Lq against OpenSim IK-derived labels; and the 2D-informed refinement is an inference-time optimization of query tokens against 2D keypoints. None of these stages defines its output in terms of its input or fits a parameter and then renames the fit as a prediction. The NeurIK losses include FK-based self-consistency terms, but these are inverse-kinematics constraints, not a tautological reduction: the target angles qr come from a separate OpenSim IK optimization, and the input markers come from image-based mesh recovery, so the learned mapping is not equal to its input by construction. The fact that the ground-truth biomechanical angles are themselves OpenSim IK outputs is a benchmark-definition choice, not a circular step, because the virtual markers are produced independently from images. The paper also does not rely on a load-bearing self-citation or an imported uniqueness theorem. The reviewer-observed discrepancy that NeurIK was trained on HMR2.0-derived markers while Table 2 reports MQ-HMR+NeurIK is a train/test distribution concern about experimental validity, not a circularity, since it does not make the reported numbers equivalent to the model's inputs. Accordingly, no circular step is identified and the score is 0.

Assumptions & free parameters 8 free parameters · 6 assumptions · 0 invented entities

The central claim rests on several hand-set hyperparameters (loss weights, token count, layer count, temporal window, iterations, marker count, camera parameters) and on the unvalidated assumption that SMPL vertices correspond to anatomical marker sites. There are no invented physical entities; the 'virtual marker' is a modeling construct whose anatomical validity is assumed rather than demonstrated.

free parameters (8)
  • MQ-HMR loss weights = λθ=1e-3, λβ=5e-4, λ3D=5e-2, λ2D=1e-2
    Hand-set in Appendix 1.2; they balance SMPL, 3D, and 2D supervision and directly affect reconstruction accuracy.
  • NeurIK loss weights = λj=1.0, λm=2.0, λs=0.1, λq=0.06
    Hand-set in Appendix 1.2; the ablation in Appendix 1.9 shows the final MAE body and MAE angle trade off with these weights.
  • Number of pose query tokens = 96
    Ablation in Section 5.1 selects 96 tokens; using 192 or 384 reduces performance.
  • Number of deformable cross-attention layers = 4
    Ablation in Appendix 1.8; 4 layers gives the best MPJPE, while 6 improves MVE slightly.
  • Temporal window length = 64 frames
    Ablation in Appendix 1.11; 64 frames gives the best MAE body and MAE angle, and 128 frames degrades results.
  • Refinement iterations = 10 (BML-MoVi, BEDLAM); 20 (OpenCap)
    Section 5.3 selects per-dataset iteration counts that minimize MAE; more iterations increase inference time.
  • Number of virtual markers = 142
    NeurIK input size given in Appendix 1.2; the vertex indices are never listed or justified.
  • Camera intrinsics = focal length 5000, identity rotation
    Appendix 1.6 fixes a weak-perspective camera; this projection model drives the 2D refinement loss and is not learned.
assumptions (6)
  • domain assumption SMPL mesh vertices can serve as virtual markers for biomechanical inverse kinematics
    Section 3.3 and Appendix 1.2; the NeurIK input is a set of 142 SMPL vertex positions, assuming these track anatomical sites the way physical skin markers do.
  • domain assumption OpenSim full-body skeleton and its joint degrees of freedom are an accurate model of human biomechanics
    Sections 3.1.2 and 3.3; qr and s ground truth are defined through OpenSim, so the skeletal model bounds the anatomical validity of all outputs.
  • ad hoc to paper The specific 142-marker vertex set is anatomically meaningful
    Appendix 1.2 gives the marker count but not the vertex indices; this choice is made by the authors and not externally validated.
  • domain assumption OpenSim Scale/IK outputs on BML-MoVi, BEDLAM, and OpenCap provide reliable ground truth
    Sections 4.1 and Appendix 1.3; all training and evaluation labels for NeurIK come from OpenSim processing of marker or mesh data.
  • domain assumption Weak-perspective camera with fixed focal length and identity rotation is sufficient for alignment
    Appendix 1.6; projection errors propagate into the 2D-informed refinement loss and 3D-to-2D consistency.
  • domain assumption 2D pose detectors such as OpenPose provide accurate keypoints for refinement
    Section 3.4; the refinement objective minimizes distance between projected 3D joints and detected 2D keypoints, so detector errors directly bias the optimized pose.

how reviews work

0 comments
Cite this review

Pith. "Pith review of BioPose: Biomechanically-accurate 3D Pose Estimation from Monocular Videos." pith.science (2026). https://pith.science/paper/X4IDAJDP

@misc{pith2026250107800,
  author       = {Pith},
  title        = {Pith review of: BioPose: Biomechanically-accurate 3D Pose Estimation from Monocular Videos},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X4IDAJDP}},
  note         = {Machine review of arXiv:2501.07800}
}
read the original abstract

Recent advancements in 3D human pose estimation from single-camera images and videos have relied on parametric models, like SMPL. However, these models oversimplify anatomical structures, limiting their accuracy in capturing true joint locations and movements, which reduces their applicability in biomechanics, healthcare, and robotics. Biomechanically accurate pose estimation, on the other hand, typically requires costly marker-based motion capture systems and optimization techniques in specialized labs. To bridge this gap, we propose BioPose, a novel learning-based framework for predicting biomechanically accurate 3D human pose directly from monocular videos. BioPose includes three key components: a Multi-Query Human Mesh Recovery model (MQ-HMR), a Neural Inverse Kinematics (NeurIK) model, and a 2D-informed pose refinement technique. MQ-HMR leverages a multi-query deformable transformer to extract multi-scale fine-grained image features, enabling precise human mesh recovery. NeurIK treats the mesh vertices as virtual markers, applying a spatial-temporal network to regress biomechanically accurate 3D poses under anatomical constraints. To further improve 3D pose estimations, a 2D-informed refinement step optimizes the query tokens during inference by aligning the 3D structure with 2D pose observations. Experiments on benchmark datasets demonstrate that BioPose significantly outperforms state-of-the-art methods. Project website: \url{https://m-usamasaleem.github.io/publication/BioPose/BioPose.html}.

Figures

Figures reproduced from arXiv: 2501.07800 by the authors.

Figure 1
Figure 1. BioPose is a framework for biomechanically accurate [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Left: Biomechanical skeleton model has anatomical de￾tails with accurate joint locations and bone orientations. Right: SMPL body model has deformable 3D body surface that includes an approximate skeleton geometry with inaccurate joint location and bone orientations. 3.1.2 Biomechanical Skeleton (BSK) Model The BSK model, e.g., widely-adopted OpenSim models, is represented by a series of bone segments that are interc… view at source ↗
Figure 3
Figure 3. Overview of BioPose, comprising two key components: (1) the [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Our 2D Pose-Guided Refinement process fine-tunes the pose query tokens to minimize 3D reconstruction errors by aligning the 3D body mesh with 2D pose estimates, improving vi￾sual coherence and providing more reliable virtual markers for the NeurIK module to further ref…
Figure 1
Figure 1. Figure 1: Comparison of state-of-the-art methods, HMR2.0 [ [PITH_FULL_IMAGE:figures/full_fig_p015_1.png]
Figure 2
Figure 2. Figure 2: Qualitative results of our approach on challenging poses from the LSP [ [PITH_FULL_IMAGE:figures/full_fig_p016_2.png]
Figure 3
Figure 3. Figure 3: Qualitative results of our proposed method BioPose and comparison with ground truth. These pictures include multiple actions [PITH_FULL_IMAGE:figures/full_fig_p017_3.png]
Figure 4
Figure 4. Figure 4: Failure Cases of MQ-HMR in 3D Human Reconstruction: MQ-HMR frequently struggles with handling unusual body movements [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Grounding Intelligence in Movement

    cs.AI 2025-07 conditional novelty 4.0 of 10

    Movement should be treated as a first-class AI modeling modality, and a unified, biomechanically grounded movement foundation model built from aggregated data across species and sensors is the proposed path forward.

Reference graph

Works this paper leans on

44 extracted references · 37 canonical work pages · cited by 1 Pith paper

  1. [1]

    Llm- assisted physical invariant extraction for cyber-physical sys- tems anomaly detection

    Danial Abshari, Chenglong Fu, and Meera Sridhar. Llm- assisted physical invariant extraction for cyber-physical sys- tems anomaly detection. arXiv preprint arXiv:2411.10918,

    arXiv

  2. [2]

    An image is worth 16x16 words: Trans- formers for image recognition at scale

    Dosovitskiy Alexey. An image is worth 16x16 words: Trans- formers for image recognition at scale. arXiv preprint arXiv: 2010.11929, 2020. 3, 4

    arXiv 2010

  3. [3]

    2d human pose estimation: New benchmark and state of the art analysis

    Mykhaylo Andriluka, Leonid Pishchulin, Peter Gehler, and Bernt Schiele. 2d human pose estimation: New benchmark and state of the art analysis. InProceedings of the IEEE Con- ference on computer Vision and Pattern Recognition , pages 3686–3693, 2014. 6

    work page 2014

  4. [4]

    Towards single camera human 3d-kinematics

    Marian Bittner, Wei-Tse Yang, Xucong Zhang, Ajay Seth, Jan van Gemert, and Frans CT van der Helm. Towards single camera human 3d-kinematics. Sensors, 23(1):341, 2022. 2, 7, 11, 12

    work page 2022

  5. [5]

    Bedlam: A synthetic dataset of bodies exhibit- ing detailed lifelike animated motion

    Michael J Black, Priyanka Patel, Joachim Tesch, and Jin- long Yang. Bedlam: A synthetic dataset of bodies exhibit- ing detailed lifelike animated motion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8726–8737, 2023. 6, 11

    work page 2023

  6. [6]

    Realtime multi-person 2d pose estimation using part affinity fields

    Zhe Cao, Tomas Simon, Shih-En Wei, and Yaser Sheikh. Realtime multi-person 2d pose estimation using part affinity fields. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 7291–7299, 2017. 2, 5, 6

    work page 2017

  7. [7]

    Video inference for human mesh recovery with vision trans- former

    Hanbyel Cho, Jaesung Ahn, Yooshin Cho, and Junmo Kim. Video inference for human mesh recovery with vision trans- former. In 2023 IEEE 17th International Conference on Automatic Face and Gesture Recognition (FG) , pages 1–6. IEEE, 2023. 3

    work page 2023

  8. [8]

    Cross- attention of disentangled modalities for 3d human mesh re- covery with transformers

    Junhyeong Cho, Kim Youwang, and Tae-Hyun Oh. Cross- attention of disentangled modalities for 3d human mesh re- covery with transformers. In European Conference on Com- puter Vision, pages 342–359. Springer, 2022. 7

    work page 2022

Show all 44 references
  1. [9]

    Beyond static features for temporally consis- tent 3d human pose and shape from a video

    Hongsuk Choi, Gyeongsik Moon, Ju Yong Chang, and Ky- oung Mu Lee. Beyond static features for temporally consis- tent 3d human pose and shape from a video. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1964–1973, 2021. 3

    1964

  2. [10]

    Learning to estimate robust 3d human mesh from in-the-wild crowded scenes

    Hongsuk Choi, Gyeongsik Moon, JoonKyu Park, and Ky- oung Mu Lee. Learning to estimate robust 3d human mesh from in-the-wild crowded scenes. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1475–1484, 2022. 3

    2022

  3. [11]

    Tokenhmr: Advancing human mesh recov- ery with a tokenized pose representation

    Sai Kumar Dwivedi, Yu Sun, Priyanka Patel, Yao Feng, and Michael J Black. Tokenhmr: Advancing human mesh recov- ery with a tokenized pose representation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1323–1333, 2024. 2, 3, 4, 5, 7, 15

    2024

  4. [12]

    Vq-hps: Hu- man pose and shape estimation in a vector-quantized latent space

    Gu ´enol´e Fiche, Simon Leglaive, Xavier Alameda-Pineda, Antonio Agudo, and Francesc Moreno-Noguer. Vq-hps: Hu- man pose and shape estimation in a vector-quantized latent space. arXiv preprint arXiv:2312.08291, 2023. 7

    2023 arXiv

  5. [13]

    Movi: A large multi-purpose human motion and video dataset

    Saeed Ghorbani, Kimia Mahdaviani, Anne Thaler, Konrad Kording, Douglas James Cook, Gunnar Blohm, and Niko- laus F Troje. Movi: A large multi-purpose human motion and video dataset. Plos one, 16(6):e0253157, 2021. 4, 6, 11

    2021

  6. [14]

    Humans in 4d: Re- constructing and tracking humans with transformers

    Shubham Goel, Georgios Pavlakos, Jathushan Rajasegaran, Angjoo Kanazawa, and Jitendra Malik. Humans in 4d: Re- constructing and tracking humans with transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 14783–14794, 2023. 2, 3, 5, 6, 7, 15

    2023

  7. [15]

    Opencapbench: A benchmark to bridge pose estimation and biomechanics

    Yoni Gozlan, Antoine Falisse, Scott Uhlrich, Anthony Gatti, Michael Black, and Akshay Chaudhari. Opencapbench: A benchmark to bridge pose estimation and biomechanics. arXiv preprint arXiv:2406.09788, 2024. 2

    2024 arXiv

  8. [16]

    Catalin Ionescu, Dragos Papava, Vlad Olaru, and Cristian Sminchisescu. Human3. 6m: Large scale datasets and pre- dictive methods for 3d human sensing in natural environ- ments. IEEE transactions on pattern analysis and machine intelligence, 36(7):1325–1339, 2013. 6

    2013

  9. [17]

    Using digital human modelling to evaluate the risk of musculoskeletal injury for workers in the health- care industry

    Xiaoxu Ji, Ranuki O Hettiarachchige, Alexa LE Littman, and Davide Piovesan. Using digital human modelling to evaluate the risk of musculoskeletal injury for workers in the health- care industry. Sensors, 23(5):2781, 2023. 1

    2023

  10. [18]

    Learning effective hu- man pose estimation from inaccurate annotation

    Sam Johnson and Mark Everingham. Learning effective hu- man pose estimation from inaccurate annotation. In CVPR 2011, pages 1465–1472. IEEE, 2011. 16

    2011

  11. [19]

    End-to-end recovery of human shape and pose

    Angjoo Kanazawa, Michael J Black, David W Jacobs, and Jitendra Malik. End-to-end recovery of human shape and pose. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7122–7131, 2018. 3

    2018

  12. [20]

    Learning 3d human dynamics from video

    Angjoo Kanazawa, Jason Y Zhang, Panna Felsen, and Jiten- dra Malik. Learning 3d human dynamics from video. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 5614–5623, 2019. 3

    2019

  13. [21]

    Emdb: The electromagnetic database of global 3d human pose and shape in the wild

    Manuel Kaufmann, Jie Song, Chen Guo, Kaiyue Shen, Tian- jian Jiang, Chengcheng Tang, Juan Jos ´e Z ´arate, and Otmar Hilliges. Emdb: The electromagnetic database of global 3d human pose and shape in the wild. In Proceedings of the IEEE/CVF International Conference on Computer ...

    2023

  14. [22]

    From skin to skeleton: Towards biomechanically accurate 3d digital hu- mans

    Marilyn Keller, Keenon Werling, Soyong Shin, Scott Delp, Sergi Pujades, C Karen Liu, and Michael J Black. From skin to skeleton: Towards biomechanically accurate 3d digital hu- mans. ACM Transactions on Graphics (TOG) , 42(6):1–12,

  15. [23]

    Pare: Part attention regressor for 3d human body estimation

    Muhammed Kocabas, Chun-Hao P Huang, Otmar Hilliges, and Michael J Black. Pare: Part attention regressor for 3d human body estimation. In Proceedings of the IEEE/CVF international conference on computer vision , pages 11127– 11137, 2021. 3, 7

    2021

  16. [24]

    Learning to reconstruct 3d human pose and shape via model-fitting in the loop

    Nikos Kolotouros, Georgios Pavlakos, Michael J Black, and Kostas Daniilidis. 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, pages 2252–2261, 2019. 6

    2019

  17. [25]

    Cliff: Carrying location information in full frames into human pose and shape estimation

    Zhihao Li, Jianzhuang Liu, Zhensong Zhang, Songcen Xu, and Youliang Yan. Cliff: Carrying location information in full frames into human pose and shape estimation. In European Conference on Computer Vision, pages 590–606. Springer, 2022. 7

    2022

  18. [26]

    One-stage 3d whole-body mesh recovery with component aware transformer

    Jing Lin, Ailing Zeng, Haoqian Wang, Lei Zhang, and Yu Li. One-stage 3d whole-body mesh recovery with component aware transformer. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 21159–21168, 2023. 4

    2023

  19. [27]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceeding...

    2014

  20. [28]

    Task- constrained motion planning considering uncertainty- informed human motion prediction for human–robot collaborative disassembly

    Wansong Liu, Xiao Liang, and Minghui Zheng. Task- constrained motion planning considering uncertainty- informed human motion prediction for human–robot collaborative disassembly. IEEE/ASME Transactions on Mechatronics, 28(4):2056–2063, 2023. 1

    2023

  21. [29]

    Matthew Loper, Naureen Mahmood, Javier Romero, Gerard Pons-Moll, and Michael J. Black. SMPL: A skinned multi- person linear model. ACM Transactions on Graphics (Proc. SIGGRAPH Asia), 34(6):248:1–248:16, oct 2015. 2, 3

    2015

  22. [30]

    3d human mesh estimation from virtual mark- ers

    Xiaoxuan Ma, Jiajun Su, Chunyu Wang, Wentao Zhu, and Yizhou Wang. 3d human mesh estimation from virtual mark- ers. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, pages 534–543, 2023. 7

    2023

  23. [31]

    Monocular 3d human pose estimation in the wild using improved cnn supervision

    Dushyant Mehta, Helge Rhodin, Dan Casas, Pascal Fua, Oleksandr Sotnychenko, Weipeng Xu, and Christian Theobalt. Monocular 3d human pose estimation in the wild using improved cnn supervision. In 2017 international con- ference on 3D vision (3DV), pages 506–516. IEEE, 2017. 6

    2017

  24. [32]

    Pose2sim: An end-to-end workflow for 3d markerless sports kinematics—part 1: Robustness

    David Pagnon, Mathieu Domalain, and Lionel Reveret. Pose2sim: An end-to-end workflow for 3d markerless sports kinematics—part 1: Robustness. Sensors, 21(19):6530,

  25. [33]

    Data augmentation for human keypoint estimation deep learning based sign lan- guage translation

    Chan-Il Park and Chae-Bong Sohn. Data augmentation for human keypoint estimation deep learning based sign lan- guage translation. Electronics, 9(8):1257, 2020. 2

    2020

  26. [34]

    Opensim: Simulating musculoskeletal dynamics and neuromuscular control to study human and animal move- ment

    Ajay Seth, Jennifer L Hicks, Thomas K Uchida, Ayman Habib, Christopher L Dembia, James J Dunne, Carmichael F Ong, Matthew S DeMers, Apoorva Rajagopal, Matthew Mil- lard, et al. Opensim: Simulating musculoskeletal dynamics and neuromuscular control to study human and animal mov...

    2018

  27. [35]

    Model-based approach for human kinematics reconstruction from marker- less and marker-based motion analysis systems

    Victor Sholukha, Bruno Bonnechere, Patrick Salvia, F ´edor Moiseev, Marcel Rooze, and S Van Sint Jan. Model-based approach for human kinematics reconstruction from marker- less and marker-based motion analysis systems. Journal of biomechanics, 46(14):2363–2371, 2013. 1

    2013

  28. [36]

    Opencap: Human movement dy- namics from smartphone videos

    Scott D Uhlrich, Antoine Falisse, Łukasz Kidzi ´nski, Julie Muccini, Michael Ko, Akshay S Chaudhari, Jennifer L Hicks, and Scott L Delp. Opencap: Human movement dy- namics from smartphone videos. PLoS computational biol- ogy, 19(10):e1011462, 2023. 1, 2, 6, 7, 11, 12

    2023

  29. [37]

    An opensim guided tour in ma- chine learning for e-health applications

    Mukul Verma, Muskan Dawar, Prashant Singh Rana, Neeru Jindal, and Harpreet Singh. An opensim guided tour in ma- chine learning for e-health applications. In Intelligent Data Security Solutions for e-Health Applications , pages 57–75. Elsevier, 2020. 1, 2

    2020

  30. [38]

    Recovering ac- curate 3d human pose in the wild using imus and a moving camera

    Timo V on Marcard, Roberto Henschel, Michael J Black, Bodo Rosenhahn, and Gerard Pons-Moll. Recovering ac- curate 3d human pose in the wild using imus and a moving camera. In Proceedings of the European conference on com- puter vision (ECCV), pages 601–617, 2018. 6

    2018

  31. [39]

    Human motion generation: A survey

    Wentao Zhu, Xiaoxuan Ma, Dongwoo Ro, Hai Ci, Jinlu Zhang, Jiaxin Shi, Feng Gao, Qi Tian, and Yizhou Wang. Human motion generation: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023. 1

    2023

  32. [40]

    Deformable detr: Deformable transform- ers for end-to-end object detection

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable detr: Deformable transform- ers for end-to-end object detection. ArXiv, abs/2010.04159,

    2010 arXiv

  33. [41]

    A comprehensive survey on the security of smart grid: Challenges, mitigations, and future research opportunities

    Arastoo Zibaeirad, Farnoosh Koleini, Shengping Bi, Tao Hou, and Tao Wang. A comprehensive survey on the security of smart grid: Challenges, mitigations, and future research opportunities. arXiv preprint arXiv:2407.07966, 2024. 1

    2024 arXiv

  34. [42]

    Vulnllmeval: A framework for evaluating large language models in soft- ware vulnerability detection and patching

    Arastoo Zibaeirad and Marco Vieira. Vulnllmeval: A framework for evaluating large language models in soft- ware vulnerability detection and patching. arXiv preprint arXiv:2409.10756, 2024. 1

    2024 arXiv

  35. [43]

    Appendix 1.1. Overview The appendix is organized into the following sections: • Section 1.2: Implementation Details • Section 1.3: Datasets • Section 1.4: Evaluation Metrics • Section 1.5: Data Augmentation • Section 1.6: Camera Model • Section 1.7: Impact of Backbones • Secti...

  36. [128]

    For the spatio-temporal model, we set the hyperparameters ex- perimentally, adjusting key parameters as needed through- out the training process

    The learning rate decayed exponentially from an initial rate of 0.001 to a final rate of 5 × 10−6 over 25 epochs. For the spatio-temporal model, we set the hyperparameters ex- perimentally, adjusting key parameters as needed through- out the training process. 1.3. Datasets We ...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.