Pith. sign in

REVIEW 4 major objections 7 minor 41 references

COMETH: Convex Optimization for Multiview Estimation and Tracking of Humans

T0 review · 4 major / 7 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read COMETH claims that modeling multi-view human pose fusion as a constrained convex inverse-kinematics problem beats two established fusion algorithms in accuracy and tracking stability on public and industrial data.

desk verdict A clearly described multi-view fusion system with a plausible method and released code, but the grid-search protocol as reported makes the headline comparison to baselines unfair without a proper train/validation/test split. read the letter →

arxiv 2508.20920 v1 pith:IC53GWI5 submitted 2025-08-28 cs.CV cs.RO

Enrico Martini , Ho Jin Choi , Nadia Figueroa , Nicola Bombieri This is my paper
classification cs.CVcs.RO
keywords multi-viewhumanposeestimation3DtrackinginversekinematicsconvexoptimizationquadraticprogrammingKalmanfilterbiomechanicalconstraintsedgecomputing
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 tries to establish that multi-camera 3D human pose tracking can be both accurate and fast enough for real-time industrial monitoring if the fusion step is built on a kinematic body model instead of generic clustering or filtering. Its central claim is that posing the fusion as a convex inverse-kinematics problem — one optimization that must satisfy every camera's keypoints under anatomical joint limits and velocity limits — produces more accurate joint positions and more stable person identities than two established fusion algorithms. On a public multi-camera benchmark the method reports the best localization, detection, association, and overall tracking scores in nearly every camera configuration, reaching 76.7 percent HOTA (the overall tracking score) with five cameras, and it keeps tracking stable with up to seven people. In a production-line laboratory it also recovers poses that a clustering-based baseline gets wrong, such as a misdetected elbow and partially occluded legs. If these results hold, edge-deployed pose tracking would no longer require trading accuracy for scalability.

What carries the argument

The central object is multi-source QPIK — quadratic-programming inverse kinematics for several simultaneous targets. At each fusion step it solves for the joint velocities of a 49-degree-of-freedom biomechanical skeleton such that the forward-kinematic joint positions match detections from every camera, with per-source slack variables that let the optimizer discount conflicting measurements. It is wrapped by two supporting mechanisms: per-bone body scaling with outlier removal, which estimates a subject's proportions before solving and discards keypoints that disagree with those proportions, and a second-order Kalman state observer that filters the 49 joint angles over time for temporal cons

What would settle it

Run the same fusion pipeline on the public benchmark but with joint-velocity limits recomputed from the evaluation subjects' own recorded motions; if the accuracy gap over the clustering baseline shrinks or disappears, the claimed gains depend on the constraint population rather than the method. A related check: count how many ground-truth benchmark poses violate the published Table 1 velocity limits when passed through the same inverse-kinematics solver.

Watch

Extended reading notes

Core claim

COMETH's central discovery is that inverse kinematics, a robotics tool, can serve as the fusion rule for multi-view human pose: rather than averaging or clustering keypoints across cameras, it treats each camera's keypoints as targets for a 49-degree-of-freedom biomechanical skeleton and solves a convex quadratic program for the joint configuration closest to all of them at once. That optimization is constrained by anatomical range-of-motion limits and by joint-velocity limits from a large motion-capture corpus, and its outputs pass through a Kalman state observer before forward kinematics reconstructs the pose. On a public benchmark it reports the highest HOTA in every camera configuration

Load-bearing premise

The joint-velocity limits are fixed from the 5th–95th percentiles of one motion-capture population, and the method assumes those same limits apply to the people in the evaluation scenes; if the deployed population moves differently, valid poses will be clipped and the reported gains will not transfer.

Editorial extensions

If this is right

  • Adding cameras steadily improves localization: reported LocA rises from 78.4% with one camera to 87.4% with five.
  • Association accuracy rises steeply with camera count, from 56.1% to 86.7%, so multi-view fusion helps identity consistency rather than hurting it.
  • Tracking remains accurate in crowded scenes: with seven subjects and five cameras, reported HOTA reaches 85.8%.
  • The fusion step is fast enough for real-time use: aggregator latency stays below the 30 Hz frame budget even as the number of incoming measurements grows.
  • The model-based constraints can correct false and missing keypoints, such as a misdetected elbow or occluded lower limbs, that clustering-based fusion propagates.

Reading between the lines

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

  • The same multi-source QPIK update could in principle be applied to any articulated body — a robot arm, an animal, or a hand — by substituting the kinematic tree and the constraint tables; nothing in the optimization itself is human-specific.
  • The paper reports only aggregate gains over baselines; ablating the three modules (per-bone scaling, convex constraints, and Kalman smoothing) separately would show which one drives the improvement.
  • Because only labeled 3D keypoints travel to the aggregator, the architecture could be extended to privacy-sensitive deployments where raw video must stay at the camera node.
  • The velocity-limit calibration should be re-estimated per deployment population: fixed 5th–95th percentile bounds from one motion-capture corpus are a prior that may need updating for worker populations with restricted mobility.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper proposes COMETH, a distributed multi-view 3D human pose fusion algorithm intended for real-time edge deployment. The pipeline performs temporal synchronization of per-camera 3D keypoints, Hungarian-based association, per-bone body scaling with outlier rejection, multi-source convex quadratic-programming inverse kinematics (QPIK) with kinematic and biomechanical joint limits, and a Kalman-filter state observer for temporal smoothing. The authors evaluate COMETH on the CMU Panoptic dataset using 1–5 RGB-D cameras against OpenPTrack and BeFine, reporting LocA, DetA, AssA, and HOTA, and they show qualitative results and latency measurements on an industrial laboratory setup. The central claim is that COMETH outperforms both baselines on tracking metrics, especially with more cameras, while remaining computationally feasible for real-time use.

Significance. If the empirical claims hold, the paper addresses a genuine engineering need: accurate multi-view human pose tracking with low communication overhead and edge-computing compatibility. The QPIK formulation with per-source slack variables, combined with a state observer and explicit biomechanical constraints, is a sensible and potentially reusable design. The release of the code and the public-dataset evaluation are commendable. However, the current validation has a load-bearing weakness: all COMETH parameters are selected by grid search, yet no train/validation/test split is documented, and the only quantitative evaluation is on the same CMU Panoptic dataset. In addition, a key equation in the state observer appears to be missing a plus sign, and Table 1 reports physically implausible velocity limits in the stated units. These issues must be resolved before the comparative claims can be accepted.

major comments (4)
  1. [§4.4 and §5.1] The paper states (Section 4.4): 'We set all COMETH parameters through grid search, keeping the same values throughout the experiments in both case studies.' The values found (Δ=0.07 s, 100 QPIK iterations, Λ=D=I, γ=1, Q=0.5·I, R=I) are then used to produce Tables 2 and 3 on the CMU Panoptic dataset. No validation split, cross-validation, or held-out sequence is described anywhere in Section 4.1 or 4.4. Since Panoptic is the only quantitative evaluation set, the grid search appears to have been performed on the same sequences used for the final reported numbers. This makes the comparison to OpenPTrack and BeFine, which are not tuned in the same way, potentially circular. The authors must either document a proper validation split, use nested cross-validation, or evaluate on sequences not touched by the grid search; otherwise the headline HOTA gains (e.g., 76.7% vs. 37.1% at five cameras) a
  2. [Table 1 and §3.3.2] Table 1 lists velocity limits in units of °/s with values such as Hip Flexion/Extension ˙qL=−1.6, ˙qU=1.9 and Knee Flexion/Extension −2.0 to 2.1. These values are three orders of magnitude too small for human joint angular velocities in degrees per second; typical hip and knee motions exceed 100°/s during walking. They are plausible only as rad/s (≈1.6 rad/s ≈ 92°/s). If the solver interprets the table as °/s, the velocity constraints in Eq. (5) would reject almost all ordinary human motion, contradicting the reported tracking performance. If the actual implementation uses rad/s, the table headers are wrong and must be corrected. Because these limits are a central component of the QPIK constraints and are taken from BioAMASS, the paper should also state the exact units used in the solver and justify their transferability to the Panoptic and ICE-Laboratory populations.
  3. [Eq. (10), §3.3.3] The Kalman-filter correction update is written as Q_i^+[t] = Q_i^-[t] · K_i[t](z_i[t] - H Q_i^-[t]). In a standard Kalman filter the correct update is Q_i^+ = Q_i^- + K_i(z_i - H Q_i^-), i.e., the innovation is added, not multiplied elementwise. As printed, Eq. (10) is not a valid state estimate: the dimensions are inconsistent and the expression does not implement the correction phase described in the text. Since the state observer is one of the three core contributions, this equation must be fixed and the implementation checked against the corrected formula.
  4. [Tables 2–3 and §5.1] The reported metrics are single-point estimates with no error bars, no standard deviations across sequences, and no significance tests. Some differences are large and likely meaningful, but others are not: for example, at two cameras the LocA values are 82.0% (COMETH) vs. 82.9% (BeFine), a difference that could easily be within run-to-run or sequence-to-sequence variability. The claim that COMETH 'significantly outperforms' the baselines is not statistically supported by the present evidence. The authors should report results per Panoptic sequence, or at least over multiple runs / subsamples, and perform a paired test where appropriate.
minor comments (7)
  1. [Throughout] The baseline name is inconsistently spelled as 'Befine' in tables and text but 'BeFine' in the abstract/references. Please unify to 'BeFine'.
  2. [Table 1] Several entries contain apparent spacing artifacts, e.g., '1 .9', '0 .5'. Please reformat the table so that all numeric values are unambiguous.
  3. [Fig. 1] In the extracted text, the region around Figure 1 contains uninterpretable glyph codes (e.g., '/gid00041/gid00084'). If this reflects the actual PDF rendering, the figure or its labels are corrupted and must be regenerated.
  4. [§3.3.2] Typo: 'mutiple' should be 'multiple'.
  5. [Eq. (1)] The notation κ2(·) is used without definition. Clarify that it selects the second-smallest element of the set of Euclidean distances, as described in the text.
  6. [§4.3] TRTPose is identified only by a footnote URL. A formal citation would be more appropriate for a comparison baseline.
  7. [§4.4] No details of the grid search are given: no ranges, no number of configurations, no criterion for selection. Even with a proper validation split, these details are needed for reproducibility.

Circularity Check

1 steps flagged · score 6.0 of 10

Panoptic HOTA/LocA comparisons are potentially fitted to the evaluation set: all COMETH parameters are grid-searched and no validation split is reported.

  1. fitted input called prediction [Section 4.1 (Datasets) and Section 4.4 (Implementation details), used in Section 5.1 Table 2]
    "We quantitatively validated the proposed fusion method on the CMU Panoptic Dataset. ... We set all COMETH parameters through grid search, keeping the same values throughout the experiments in both case studies."

    The only quantitative evaluation is the CMU Panoptic dataset, yet no train/validation/test split is described. The grid search over Δ, QPIK iterations, Λ, D, γ, Q, and R is therefore indistinguishable from selecting the configuration that maximizes LocA/DetA/AssA/HOTA on the same sequences that produce Table 2. The reported gains (e.g., HOTA 76.7% vs BeFine's 37.1% at five cameras) are the outcome of an optimization on the evaluation set, not an independent prediction. This makes the headline comparison statistically forced rather than a fair algorithmic test.

full rationale

The algorithmic derivation of COMETH is not circular: QPIK, BSM constraints, Kalman smoothing, and BioAMASS velocity bounds are external inputs, and the height-scaling citation [30] plus the BeFine baseline [10] are self-citations that do not by themselves force the result. The load-bearing circularity is in the evaluation protocol. Section 4.4 states all parameters are set by grid search, and Section 4.1 identifies the CMU Panoptic dataset as the quantitative validation set, with no reported split. Thus Table 2's numbers are at risk of being fitting results rather than predictions. The ICE Laboratory results are qualitative and cannot independently validate the quantitative claims. Score 6 reflects that the central empirical claim partially reduces to a fit; it is not 8 or 10 because the method's internal optimization is still a genuine algorithm and the BioAMASS/position-limit inputs are not renamed predictions.

Assumptions & free parameters 10 free parameters · 6 assumptions · 0 invented entities

The central claim rests on the BSM body model, anthropometric height proportions, local linearization in QPIK, percentile-fitted velocity limits, a constant-acceleration Kalman model, and the accuracy of depth-based 3D keypoints. The largest unverified loads are the transferability of BioAMASS-derived bounds and the grid-searched parameters to the evaluation data.

free parameters (10)
  • Window timeout Delta = 0.07 s
    Measurements older than Delta are discarded; set by grid search on the Panoptic dataset (Section 4.4).
  • QPIK scaling factor gamma = 1
    Controls how fast joints approach their range-of-motion limits; grid-searched and fixed for all experiments.
  • QPIK weight matrices Lambda and D = identity matrices
    Weights in the quadratic objective; chosen by hand or grid search, not derived from data.
  • Kalman process noise covariance Q/Sigma = 0.5 I
    Set by grid search in Section 4.4; not estimated from sensor statistics.
  • Kalman measurement noise covariance R = I
    Set by grid search in Section 4.4.
  • Maximum QPIK iterations = 100
    Stopping criterion chosen in implementation; no sensitivity analysis reported.
  • Per-bone scaling clip width = plus or minus 5% of average scale
    Body part scales are clipped to stay near the average height-derived scale; 5% is a hand-chosen tolerance (Section 3.3.1).
  • Joint velocity bounds = 5th and 95th percentiles of BioAMASS joint velocities
    Velocity limits in Table 1 are fitted from the BioAMASS dataset and assumed to generalize to evaluation populations.
  • Joint range-of-motion bounds = normative values from literature, e.g., elbow -11 to 154 degrees
    Anatomical range limits are taken from biomechanics references and treated as fixed.
  • Minimum valid keypoint count k = not reported
    Measurements require at least k valid keypoints before association (Section 3.2); the value of k is not given.
assumptions (6)
  • domain assumption The human body is represented by the BSM biomechanical model with 49 DOFs and 24 rigid bone links.
    Adopted from Keller et al. [24]; COMETH's inverse kinematics and forward kinematics depend on this model's topology and joint definitions (Section 3.2).
  • domain assumption Human segment lengths are proportional to total height according to anthropometric tables.
    Body scaling estimates height from inter-joint distances using proportions from [28,29] (Section 3.3.1).
  • standard math First-order Taylor linearization x + J(q) dq = target captures the kinematics within each QP iteration.
    The QPIK constraint (Equation 3) relies on iterative local linearization; convergence is assumed.
  • ad hoc to paper The 5th and 95th percentile velocity bounds from BioAMASS are valid for all test subjects.
    There is no independent evidence that these percentile limits transfer to CMU Panoptic actors or factory workers (Table 1).
  • standard math The second-order constant-acceleration model F describes human joint motion between frames.
    The Kalman filter prediction model (Equation 6) assumes constant acceleration per joint in a 30 Hz stream.
  • domain assumption Depth-based back-projection yields sufficiently accurate 3D keypoints for meaningful fusion.
    All methods, including baselines, consume 3D keypoints from TRTPose or OpenPose plus depth; errors in this input propagate to all outputs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of COMETH: Convex Optimization for Multiview Estimation and Tracking of Humans." pith.science (2026). https://pith.science/paper/IC53GWI5

@misc{pith2026250820920,
  author       = {Pith},
  title        = {Pith review of: COMETH: Convex Optimization for Multiview Estimation and Tracking of Humans},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IC53GWI5}},
  note         = {Machine review of arXiv:2508.20920}
}
read the original abstract

In the era of Industry 5.0, monitoring human activity is essential for ensuring both ergonomic safety and overall well-being. While multi-camera centralized setups improve pose estimation accuracy, they often suffer from high computational costs and bandwidth requirements, limiting scalability and real-time applicability. Distributing processing across edge devices can reduce network bandwidth and computational load. On the other hand, the constrained resources of edge devices lead to accuracy degradation, and the distribution of computation leads to temporal and spatial inconsistencies. We address this challenge by proposing COMETH (Convex Optimization for Multiview Estimation and Tracking of Humans), a lightweight algorithm for real-time multi-view human pose fusion that relies on three concepts: it integrates kinematic and biomechanical constraints to increase the joint positioning accuracy; it employs convex optimization-based inverse kinematics for spatial fusion; and it implements a state observer to improve temporal consistency. We evaluate COMETH on both public and industrial datasets, where it outperforms state-of-the-art methods in localization, detection, and tracking accuracy. The proposed fusion pipeline enables accurate and scalable human motion tracking, making it well-suited for industrial and safety-critical applications. The code is publicly available at https://github.com/PARCO-LAB/COMETH.

Figures

Figures reproduced from arXiv: 2508.20920 by the authors.

Figure 1
Figure 1. Overview of the multi-camera fusion pipeline. Edge devices independently perform HPE from multiple camera views. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Graphical representation of the body scaling and [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Results on the ICE Laboratory setup from the proposed fusion method. The first three columns are the synchronized [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison of the fusion methods for multiple moving subjects on the real-factory scenario, along with the ground [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Mean latency of the compared fusion methods [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

41 extracted references · 37 canonical work pages

  1. [1]

    Lupi´ on, A

    M. Lupi´ on, A. Polo-Rodr ´ ıguez, J. Medina-Quero, J. F. Sanjuan, P. M. Ortigosa, 3D Human Pose Estimation from multi-view thermal vision sensors, Inf. Fusion 104 (November 2023) (2024) 102154. doi:10.1016/j.inffus.2023.102154. URL https://doi.org/10.1016/j.inffus.2023. 102154 1

    work page arXiv 2023

  2. [2]

    Geretti, S

    L. Geretti, S. Centomo, M. Boldo, E. Martini, N. Bombieri, D. Quaglia, T. Villa, Process-driven colli- sion prediction in human-robot work environments, in: 2022 IEEE 27th International Conference on Emerging Technologies and Factory Automation (ETF A), IEEE, 2022, pp. 1–8. 1

    work page 2022

  3. [3]

    Rathor, S

    K. Rathor, S. Lenka, K. A. Pandya, B. Gokulakrishna, S. S. Ananthan, Z. T. Khan, A detailed view on indus- trial safety and health analytics using machine learn- ing hybrid ensemble techniques, in: 2022 International Conference on Edge Computing and Applications (ICE- CAA), IEEE, 2022, pp. 1166–1169. 1

    work page 2022

  4. [4]

    R. Bazo, E. Reis, L. A. Seewald, V. F. Rodrigues, C. A. da Costa, L. Gonzaga, R. S. Antunes, R. d. R. Righi, A. Maier, B. Eskofier, R. Fahrig, T. Horz, Baptizo: A sensor fusion based model for tracking the identity of human poses, Inf. Fusion 62 (February) (2020) 1–13. doi:10.1016/j.inffus.2020.03.011. 1

    work page doi:10.1016/j.inffus.2020.03.011 2020

  5. [5]

    Zhang, C

    Y. Zhang, C. Jiang, B. Yue, J. Wan, M. Guizani, Infor- mation fusion for edge intelligence: A survey, Informa- tion Fusion 81 (2022) 171–186. 1

    work page 2022

  6. [6]

    C. Xu, J. He, X. Zhang, C. Yao, P. H. Tseng, Geometri- cal kinematic modeling on human motion using method of multi-sensor fusion, Inf. Fusion 41 (2018) 243–254. doi:10.1016/j.inffus.2017.09.014. 1, 3

    work page doi:10.1016/j.inffus.2017.09.014 2018

  7. [7]

    Real-time marker-less multi-person 3D pose estimation in RGB-Depth camera networks

    M. Carraro, M. Munaro, J. Burke, E. Menegatti, Real- time marker-less multi-person 3D pose estimation in 11 RGB-DEPTH camera networks, Adv. Intell. Syst. Com- put. 867 (2019) 534–545. arXiv:1710.06235, doi:10. 1007/978-3-030-01370-7_42 . 1, 8, 9, 10

    work page Pith review arXiv 2019

  8. [8]

    E. A. Wan, R. Van Der Merwe, The unscented kalman filter for nonlinear estimation, in: Proceedings of the IEEE 2000 adaptive systems for signal process- ing, communications, and control symposium (Cat. No. 00EX373), Ieee, 2000, pp. 153–158. 1

    work page 2000

Show all 41 references
  1. [9]

    Z. Cao, G. Hidalgo Martinez, T. Simon, S. Wei, Y. A. Sheikh, Openpose: Realtime multi-person 2d pose esti- mation using part affinity fields, IEEE Transactions on Pattern Analysis and Machine Intelligence (2019). 2, 8

    2019

  2. [10]

    Boldo, M

    M. Boldo, M. De Marchi, E. Martini, S. Aldegheri, D. Quaglia, F. Fummi, N. Bombieri, Real-time multi- camera 3D human pose estimation at the edge for indus- trial applications, Expert Syst. Appl. 252 (PA) (2024) 124089. doi:10.1016/j.eswa.2024.124089. URL https://doi.org/10.10...

    2024

  3. [11]

    H. Joo, T. Simon, X. Li, H. Liu, L. Tan, L. Gui, S. Banerjee, T. S. Godisart, B. Nabbe, I. Matthews, T. Kanade, S. Nobuhara, Y. Sheikh, Panoptic studio: A massively multiview system for social interaction cap- ture, IEEE Transactions on Pattern Analysis and Ma- chine Intellige...

    2017

  4. [12]

    Nexus, Vicon nexus product guide (2015)

    V. Nexus, Vicon nexus product guide (2015). URL https://documentation.vicon.com/nexus/v2.2/ Nexus1_8Guide.pdf 2

    2015

  5. [13]

    Desmarais, D

    Y. Desmarais, D. Mottet, P. Slangen, P. Montesinos, A review of 3d human pose estimation algorithms for markerless motion capture, Computer Vision and Image Understanding 212 (2021) 103275. 2

    2021

  6. [14]

    Sarafianos, B

    N. Sarafianos, B. Boteanu, B. Ionescu, I. A. Kakadiaris, 3d human pose estimation: A review of the literature and analysis of covariates, Computer Vision and Image Understanding 152 (2016) 1–20. 2

    2016

  7. [15]

    T. B. Moeslund, A. Hilton, V. Kr¨ uger, A survey of ad- vances in vision-based human motion capture and anal- ysis, Computer vision and image understanding 104 (2-

  8. [16]

    Zheng, W

    C. Zheng, W. Wu, C. Chen, T. Yang, S. Zhu, J. Shen, N. Kehtarnavaz, M. Shah, Deep learning-based human pose estimation: A survey, ACM Computing Surveys 56 (1) (2023) 1–37. 2

    2023

  9. [17]

    Martini, A

    E. Martini, A. Calanca, N. Bombieri, Denoising and completion filters for human motion software: a survey with code, Authorea Preprints (2023). 2

    2023

  10. [18]

    Zhang, J

    Y. Zhang, J. Zhang, S. Xu, J. Xiao, Multi-view human pose and shape estimation via mesh-aligned voxel in- terpolation, Information Fusion 114 (2025) 102651. 2

    2025

  11. [19]

    Denavit, R

    J. Denavit, R. S. Hartenberg, A kinematic notation for lower-pair mechanisms based on matrices (1955). 2

    1955

  12. [20]

    Koptev, N

    M. Koptev, N. Figueroa, A. Billard, Real-Time Self- Collision Avoidance in Joint Space for Humanoid Robots, IEEE Robot. Autom. Lett. 6 (2) (2021) 1240–

    2021

  13. [21]

    Loper, N

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

    2015

  14. [22]

    S. L. Delp, F. C. Anderson, A. S. Arnold, P. Loan, A. Habib, C. T. John, E. Guendelman, D. G. Thelen, Opensim: open-source software to create and analyze dynamic simulations of movement, IEEE transactions on biomedical engineering 54 (11) (2007) 1940–1950. 3

    2007

  15. [23]

    Werling, J

    K. Werling, J. Kaneda, A. Tan, R. Agarwal, S. Skov, T. Van Wouwe, S. Uhlrich, N. Bianco, C. Ong, A. Falisse, S. Sapkota, A. Chandra, J. Carter, E. Preatoni, B. Fregly, J. Hicks, S. Delp, C. K. Liu, Addbiomechanics dataset: Capturing the physics of human motion at scale (2024)....

    2024 doi arXiv

  16. [24]

    Keller, K

    M. Keller, K. Werling, S. Shin, S. Delp, S. Pujades, C. K. Liu, M. J. Black, From skin to skeleton: To- wards biomechanically accurate 3D digital humans, ACM Transaction on Graphics (ToG) 42 (6) (2023) 253:1–253:15. doi:https://doi.org/10.1145/3618381. 3, 4, 6

    2023 doi

  17. [25]

    Mahmood, N

    N. Mahmood, N. Ghorbani, N. F. Troje, G. Pons-Moll, M. J. Black, Amass: Archive of motion capture as sur- face shapes, in: Proceedings of the IEEE/CVF interna- tional conference on computer vision, 2019, pp. 5442–

    2019

  18. [26]

    Siciliano, O

    B. Siciliano, O. Khatib, T. Kr¨ oger, Springer handbook of robotics, Vol. 200, Springer, 2008. 3

    2008

  19. [27]

    S. P. Boyd, L. Vandenberghe, Convex optimization, Cambridge university press, 2004. 3, 6

    2004

  20. [28]

    Drillis, R

    R. Drillis, R. Contini, M. Bluestein, Body segment pa- rameters, Artificial limbs 8 (1) (1964) 44–66. 5

    1964

  21. [29]

    D. A. Winter, Biomechanics and motor control of hu- man movement, John wiley & sons, 2009. 5

    2009

  22. [30]

    Martini, H

    E. Martini, H. Parekh, S. Peng, N. Bombieri, N. Figueroa, A robust filter for marker-less multi-person tracking in human-robot interaction scenarios, in: 2024 33rd IEEE International Conference on Robot and Human Interactive Communication (ROMAN), IEEE, 2024, pp. 424–429. 5

    2024

  23. [31]

    E. L. Zwerus, N. W. Willigenburg, V. A. Scholtes, M. P. Somford, D. Eygendaal, M. P. van den Bekerom, Nor- mative values and affecting factors for the elbow range of motion, Shoulder & elbow 11 (3) (2019) 215–224. 5

    2019

  24. [32]

    A. Ames, S. D. Coogan, M. Egerstedt, G. Notomista, K. Sreenath, P. Tabuada, Control barrier functions: Theory and applications, 2019 18th European Control Conference (ECC) (2019) 3420–3431. 5

    2019

  25. [33]

    Levenberg, A method for the solution of certain non- linear problems in least squares, Quarterly of applied mathematics 2 (2) (1944) 164–168

    K. Levenberg, A method for the solution of certain non- linear problems in least squares, Quarterly of applied mathematics 2 (2) (1944) 164–168. 6

    1944

  26. [34]

    R. E. Kalman, A new approach to linear filtering and prediction problems (1960). 6

    1960

  27. [35]

    Cunico, S

    F. Cunico, S. Aldegheri, A. Avogaro, M. Boldo, N. Bombieri, L. Capogrosso, A. Caputo, D. Carra, S. Centomo, D. S. Cheng, E. Cinquetti, M. Cristani, M. D. Marchi, F. Demrozi, M. Emporio, F. Fummi, L. Geretti, S. Germiniani, A. Giachetti, F. Girella, E. Martini, G. Menegaz, N. M...

    2024

  28. [36]

    Luiten, A

    J. Luiten, A. Osep, P. Dendorfer, P. Torr, A. Geiger, L. Leal-Taix´ e, B. Leibe, HOTA: A Higher Order Metric for Evaluating Multi-object Tracking, Int. J. Comput. Vis. 129 (2) (2021) 548–578. arXiv:2009.07736, doi: 10.1007/s11263-020-01375-2 . 7

    2021 arXiv

  29. [37]

    H. W. Kuhn, The hungarian method for the assign- ment problem, Naval research logistics quarterly 2 (1-2) 12 (1955) 83–97. 7

    1955

  30. [38]

    Werling, D

    K. Werling, D. Omens, J. Lee, I. Exarchos, K. Liu, Fast and feature-complete differentiable physics engine for articulated rigid bodies with contact constraints, in: Robotics: Science and Systems XVII, RSS2021, Robotics: Science and Systems Foundation, 2021. doi:10.15607/rss.2...

    2021 doi

  31. [39]

    J. Lee, M. X. Grey, S. Ha, T. Kunz, S. Jain, Y. Ye, S. S. Srinivasa, M. Stilman, C. Karen Liu, Dart: Dy- namic animation and robotics toolkit, The Journal of Open Source Software 3 (22) (2018) 500.doi:10.21105/ joss.00500. URL http://dx.doi.org/10.21105/joss.00500 8

    2018 doi

  32. [40]

    Diamond, S

    S. Diamond, S. Boyd, Cvxpy: A python-embedded modeling language for convex optimization, Journal of Machine Learning Research 17 (83) (2016) 1–5. 8 13

    2016

  33. [1247]

    doi:10.1109/LRA.2021.3057024. 2, 3

    2021

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.