Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

A two-stage optimizer that adds whole-body dynamics and ground-reaction-force heel–toe contacts to motion retargeting produces dynamically feasible humanoid references and measurably faster imitation learning.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review

2026-08-03 00:51 UTC pith:YBMURYOI

load-bearing objection Promising kinodynamic retargeting pipeline, but as written the dynamics constraints in Eqs. 10–11 are wrong and the empirical evidence is under-quantified. the 4 major comments →

arxiv 2603.09956 v2 pith:YBMURYOI submitted 2026-03-10 cs.RO

Kinodynamic Motion Retargeting for Humanoid Locomotion via Multi-Contact Whole-Body Trajectory Optimization

Xiaoyu Zhang , Steven Haener , Varun Madabushi , Maegan Tucker This is my paper
classification cs.RO
keywords kinodynamic motion retargetinghumanoid locomotionwhole-body trajectory optimizationground reaction forcesmulti-contactimitation learningheel-toe rollcontact complementarity
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that fitting human motion-capture data onto a humanoid robot by kinematics alone creates physically impossible references that slow imitation learning. It introduces KDMR, a kinodynamic retargeting pipeline that starts with a kinematic fit, then refines the trajectory through a whole-body optimization enforcing rigid-body dynamics, contact complementarity, friction-cone bounds, and trapezoidal integration. Crucially, KDMR uses ground reaction force (GRF) data to label heel-only, flat-foot, and toe-only contact windows, reproducing the natural heel-to-toe roll. The authors claim this yields trajectories without foot sliding or ground penetration, and that policies trained on them converge faster and reach higher reward than those trained on a purely kinematic baseline.

Core claim

KDMR establishes that integrating GRF-derived contact patterns and explicit multi-contact dynamics into motion retargeting produces dynamically viable reference trajectories that improve downstream imitation learning. The dynamic optimization jointly solves for joint positions, velocities, torques, and heel/toe contact forces subject to the floating-base rigid-body dynamics, zero-velocity and zero-acceleration constraints at active contacts, friction pyramid bounds, and trapezoidal integration, while a weighted cost tracks the kinematic reference, smooths velocities and forces, and penalizes ground penetration. Compared with a purely kinematic retargeting baseline, KDMR removes foot floating

What carries the argument

The load-bearing mechanism is the heel-to-toe roll contact decomposition drawn from the double-peak vertical GRF waveform: a mid-stance dip marks the flat-foot phase with both heel and toe in contact, and the preceding and succeeding non-zero windows are labeled heel-only and toe-only. These binary contact labels drive a multi-contact whole-body trajectory optimization that combines a kinematic initialization (differential inverse kinematics) with a nonlinear dynamic refinement enforcing floating-base dynamics, holonomic contact constraints, a linearized friction pyramid, and trapezoidal integration, with a cost function that penalizes base and joint tracking error, velocity and force discon

Load-bearing premise

The contact labels (heel-only, flat-foot, toe-only) are extracted from a steady-state walking GRF profile; if the human motion is not steady-state walking, or if GRF data is intermittent, the contact sequence fed into the dynamic optimizer is wrong and the entire dynamically feasible retargeting rests on incorrect ground contacts.

What would settle it

Retarget a non-steady-state motion (a turn, a stop, a step up, or running) from a similar dataset, or use GRF data interrupted by overground force-plate boundaries, and check whether KDMR still yields contact-consistent, slip-free trajectories; any failure would show the claim holds only for steady-state walking with continuous synchronized GRF.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Imitation-learned humanoid policies can be trained on dynamically feasible references, so the reinforcement learning agent no longer spends sample budget compensating for impossible motions like foot penetration or base floating.
  • The GRF-based contact detector may enable natural heel-to-toe roll on humanoids without manual contact scheduling, improving gait stability and stride efficiency.
  • The two-stage approach—kinematic warm-start followed by dynamic refinement—is a transferable template for other retargeting tasks where contact sequences are known or can be estimated.
  • Because the dynamic optimization is motion-agnostic once contact sequences are supplied, the method could extend to arbitrary gaits provided a reliable contact estimator exists.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The contact-extraction rule is tied to steady-state walking with a characteristic double-peak GRF; a natural stress test is whether the pipeline works on running, turning, stair ascent, or intermittent overground force-plate data, where the waveform differs—if not, the contribution is scoped to walking.
  • Rescaling GRF by mass ratio and modeling heel and toe as three-dimensional point contacts could make the method transferable across humanoid platforms, but the point-contact assumption may not capture rolling effects on feet with compliant soles.
  • The reported sample-efficiency gain suggests that many prior imitation results may have been limited as much by reference trajectory quality as by policy architecture or reward design—an implication worth testing on other benchmark motions.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes KDMR, a kinodynamic motion-retargeting pipeline for humanoid locomotion. Human MoCap marker data are converted to human skeleton poses with OpenSim, then to an initial robot reference via differential IK. A heuristic based on the vertical GRF double-peak profile labels heel/flat/toe contact phases and splits the total GRF into heel and toe components. A dynamic trajectory-optimization stage then adjusts the reference using whole-body rigid-body dynamics, contact and friction constraints, and a trapezoidal integration law, producing a dynamically feasible multi-contact reference. The method is evaluated on Unitree G1 with a walking dataset, comparing against the kinematic GMR baseline, and the retargeted references are used to train BeyondMimic policies to assess downstream sample efficiency.

Significance. If the formulation and experiments are correct, the paper addresses an important gap in humanoid motion retargeting: purely kinematic references commonly contain foot sliding, penetration, and dynamically infeasible transitions, which hurt imitation learning. Explicitly embedding heel-toe multi-contact dynamics and GRF-derived contact states into the optimization is a sensible and timely idea. The paper also has concrete strengths: the problem formulation is clearly structured, the downstream policy comparison uses five seeds, and the authors state an intention to open-source the pipeline. However, the current manuscript has a serious error in the integration constraints that directly affects the dynamic-feasibility claim, and the experimental sections rely mostly on qualitative visual comparisons rather than quantitative metrics. For these reasons, the paper needs substantial revision before the claims can be accepted.

major comments (4)
  1. [IV-B, Figs. 5-7] The 'trapezoidal integration law' is written as q_{t+1}=q_t + (Δt/2)(qdot_{t+1} - qdot_t) and similarly for velocity. This is not the trapezoidal rule; the correct form uses the sum of the endpoint derivatives: q_{t+1}=q_t + (Δt/2)(qdot_t + qdot_{t+1}). As printed, a constant velocity produces q_{t+1}=q_t, so the constraints do not enforce any relation between position, velocity, and acceleration. Since the central claim of dynamic feasibility in §III-E and the Abstract rests on these constraints, this is a load-bearing issue. Please correct the equations, confirm that the implementation follows the corrected form, and add a numerical consistency check (e.g., forward simulation of the optimized trajectory or reported integration residuals). If this was only a typographical error, state so explicitly.
  2. [IV-B, Fig. 6b] The experimental evaluation is qualitative. Fig. 5 and Fig. 6a report time traces but give no quantitative metrics for foot sliding, ground penetration, foot-height error, or smoothness; Fig. 7 shows mean reward curves with a shaded region but no numerical values, effect sizes, or statistical tests. The Abstract claims KDMR 'significantly outperforms' GMR. Such claims need numbers: e.g., mean ± std of penetration depth, foot sliding distance, reward after a fixed number of steps, and confidence intervals over the five seeds. Please add a quantitative summary table for the artifact metrics and the policy comparison.
  3. [IV-C] The GRF-tracking evaluation is partly circular. The heel/toe contact windows used in the optimization (§III-C) are derived from the same vertical GRF source that is used as the 'Source' curves in Fig. 6b, and the total force is split into heel and toe components using the same heuristic. Matching that decomposition is therefore partly by construction. Please validate against the raw measured total GRF (and ideally independent foot-pressure data), or at minimum quantify and discuss how much of the agreement is attributable to the contact-state labeling heuristic.
  4. [III-E] The downstream comparison includes only the purely kinematic GMR baseline. While this is appropriate for showing improvement over kinematic retargeting, the paper's framing as a 'kinodynamic' contribution would be strengthened by a comparison with at least one dynamics-aware retargeting baseline, such as DynaRetarget [24] or a simplified dynamics filter. Without such a comparison, the reader cannot tell whether the reported policy improvements come from the specific multi-contact dynamics formulation or simply from any dynamics-aware correction.
minor comments (6)
  1. [III-E, cost function] In the displayed cost function, the plus signs between the force-smoothing term, torque regularization, and ground-contact term are missing, making the expression ambiguous. Please insert them.
  2. [III-C] The heel/toe contact-window extraction relies on thresholds such as 'periods of non-zero vertical force' and the mid-stance dip. Please specify the exact criteria, thresholds, and smoothing used, since this directly affects reproducibility and the contact labels fed to the optimizer.
  3. [I, contribution 1] The phrase 'mathematically guaranteed, smooth and dynamically feasible reference trajectories' overstates what the optimization delivers: the ground-contact constraint (6) is included as a soft cost, the dynamics are discretized, and the integration constraints are only as correct as Eqs. (10)-(11). Please temper the claim to reflect the actual formulation.
  4. [IV-B, Fig. 5] The caption says 'foot pitch (φ_z)' and the text mentions 'foot ankle pitch (θ_4)', but the joint naming is inconsistent. Please define the joint ordering and the meaning of θ_Li once in the text.
  5. [IV-C, Fig. 7] The moving-average filter used for visual clarity is not described. Please state the window length and whether the statistics are computed before or after filtering.
  6. [V] The limitation section acknowledges the steady-state walking assumption and the need for synchronized motion-force data, but it does not mention the integration-constraint issue or the circularity in the GRF-tracking evaluation. Please extend the limitations accordingly after addressing the major comments.

Circularity Check

1 steps flagged

Central policy/dynamics claim is independently benchmarked, but the contact-timing 'recovery' evaluation is circular: the optimizer is handed the GRF-derived contact windows and then credited with recovering them.

specific steps
  1. fitted input called prediction [Section III-C -> Section IV-B (Fig. 6b); Conclusion]
    "we deduce the discrete contact states for the heel and toe directly from the vertical GRF profile of the corresponding foot. Specifically, we identify a temporal window centered around the mid-stance dip to represent the flat-foot phase, labeling both the heel and toe as being in contact. The periods of non-zero vertical force preceding and succeeding this window are strictly extracted as heel-only and toe-only contacts, respectively. ... With the multi-contact formulation, our method successfully recovers both the timing of the foot contact events and the magnitude of the required forces."

    The binary contact sequence (heel-only / flat / toe-only) is not a prediction of the optimizer: Section III-C constructs it from the source vertical GRF profile, and the dynamic NLP then hard-constrains the trajectory to respect exactly those labels through Eqs. (4)-(6), (7)-(9), with stationarity and zero height during the same windows. Thus the retargeted trajectory's contact timing is forced to match the source-GRF windows by construction. Section IV-B presenting this as 'recovers the timing of the foot contact events' compares the output to its own input; it cannot independently validate that KDMR extracts contact timing from GRF. The GRF magnitude comparison is less directly forced because no lambda-tracking term appears in the cost, so the circularity is specific to the timing sub-cl

full rationale

The central derivation is not circular. The dynamic optimization takes the kinematic retarget q_ref and the contact labels as inputs and solves an NLP; the downstream policy comparison trains BeyondMimic policies on both KDMR and GMR references under identical hyperparameters, so the sample-efficiency conclusion is an independent empirical benchmark. There is no load-bearing self-citation chain: all cited dynamics/GRF results are external, and no uniqueness theorem is invoked. The one genuinely circular sub-claim is the contact-event timing evaluation: III-C constructs heel/toe contact windows directly from the vertical GRF profile and feeds them into the optimizer as hard contact constraints; IV-B then treats the optimizer's adherence to those windows as 'recovering the timing of foot contact events' from source data. That specific evaluation reduces to its own input by construction. The magnitude-tracking comparison is less obviously forced, so it is not scored as independently circular. The trapezoid-integration sign issue in Eqs. (10)-(11) is a correctness/implementation defect, not a circularity, and section V's stated limitation about steady-state walking is an acknowledged assumption rather than a circular step. Overall, the paper retains independent content in its main claim, hence a moderate score rather than a higher one.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

No new physical entities are introduced. Heel/toe contact domains are biomechanical modeling constructs, not invented entities. The main free parameters are hand-tuned cost weights and unspecified contact-window thresholds, plus standard domain assumptions about gait, rigid-body modeling, and data synchronization.

free parameters (2)
  • Cost weights w_b, w_λ, w_j, w_v, w_τ, w_c = w_b=diag([10,10,3,5,5,5]), w_λ=1e-4, w_j=diag([2,5,5,2,2,2]) per side, w_v=1e-3, w_τ=1e-4, w_c=1e3
    Empirically configured in Section IV-A without a tuning procedure or ablation; optimal trajectories and constraint satisfaction depend on these values.
  • Heel/toe contact-window thresholds = not reported
    Section III-C extracts a flat-foot window around the mid-stance dip; the width and thresholds of this window are not specified, yet they determine which contacts are enforced in Equations (3)-(9).
axioms (4)
  • domain assumption Human steady-state walking has a double-peak vertical GRF profile from which heel-only, flat-foot, and toe-only phases can be segmented.
    Section III-C uses this to label contact states; Section V explicitly acknowledges the assumption is specific to steady-state walking.
  • domain assumption Rigid-body dynamics (Equation 3) and a linearized friction pyramid (Equations 7-9) accurately model the Unitree G1 robot and its floor contacts.
    Standard modeling assumptions, but the point-contact approximation neglects rotational contact wrenches (Section III-E), and no hardware validation is provided.
  • domain assumption The OpenSim inverse kinematics step maps marker trajectories to human skeleton poses accurately.
    Section III-B relies on OpenSim; errors in marker-to-skeleton fitting propagate to all target poses and hence to the retargeting.
  • standard math The trapezoidal integration law as printed (Equations 10-11, with minus signs) is a valid discrete-time integration scheme.
    Standard trapezoidal integration uses a plus sign; as printed the equations do not correctly relate q, q_dot, and q_ddot, which would invalidate the NLP. Likely a typo, but it must be corrected.

pith-pipeline@v1.3.0-alltime-deepseek · 10708 in / 9828 out tokens · 92274 ms · 2026-08-03T00:51:57.814594+00:00 · methodology

0 comments
read the original abstract

We present the KinoDynamic Motion Retargeting (KDMR) framework, a novel approach for humanoid locomotion that models the retargeting process as a multi-contact, whole-body trajectory optimization problem. Conventional kinematics-based retargeting methods rely solely on spatial motion capture (MoCap) data, inevitably introducing physically inconsistent artifacts, such as foot sliding and ground penetration, that severely degrade the performance of downstream imitation learning policies. To bridge this gap, KDMR extends beyond pure kinematics by explicitly enforcing rigid-body dynamics and contact complementarity constraints. Further, by integrating ground reaction force (GRF) measurements alongside MoCap data, our method automatically detects heel-toe contact events to accurately replicate complex human-like contact patterns. We evaluate KDMR against the state-of-the-art baseline, GMR, across three key dimensions: 1) the dynamic feasibility and smoothness of the retargeted motions, 2) the accuracy of GRF tracking compared to raw source data, and 3) the training efficiency and final performance of downstream control policies trained via the BeyondMimic framework. Experimental results demonstrate that KDMR significantly outperforms purely kinematic methods, yielding dynamically viable reference trajectories that accelerate policy convergence and enhance overall locomotion stability. Our end-to-end pipeline will be open-sourced upon publication.

Figures

Figures reproduced from arXiv: 2603.09956 by Maegan Tucker, Steven Haener, Varun Madabushi, Xiaoyu Zhang.

Figure 1
Figure 1. Figure 1: The presented Kinodynamic Motion Retargeting (KDMR) framework translates human motion references and ground reaction forces into retargeted [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Natural human walking, which we call multi-contact locomotion, is characterized by a heel-to-toe rolling motion. The diagram illustrates one complete gait cycle, beginning and ending with right heel strike. multi-contact locomotion. Our approach processes human motion capture data by leveraging ground reaction forces as a primary physical anchor. By formulating the retargeting process to account for the co… view at source ↗
Figure 3
Figure 3. Figure 3: Algorithm overview: The KDMR algorithm converts time-series marker and ground reaction force data into joint position targets for the humanoid [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Illustration of the human GRF pattern used to estimate the contact [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of retargeted trajectories between KDMR and GMR over a selected time segment (32–40 s) from the full motion sequence. (a) Robot [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of heel–toe height and contact forces with respect to the baseline method. (a) Heel and toe heights of the left foot for KDMR and [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Comparison of mean episode reward and value function loss over five independent training runs using BeyondMimic for each method. All methods [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

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

  1. Booster Lab: A Data-Centric Pipeline for Learning Deployable Humanoid Locomotion Policies

    cs.RO 2026-06 unverdicted novelty 4.0

    Describes an integrated pipeline for curating motion data, adapting real-to-sim models, applying AMP-based RL, and deploying locomotion policies on Booster T1 and K1 humanoid robots.

Reference graph

Works this paper leans on

34 extracted references · 11 linked inside Pith · cited by 1 Pith paper

  1. [1]

    AMP: Adversarial motion priors for stylized physics-based character con- trol,

    X. B. Peng, Z. Ma, P. Abbeel, S. Levine, and A. Kanazawa, “AMP: Adversarial motion priors for stylized physics-based character con- trol,”ACM Transactions on Graphics (ToG), 2021

    2021

  2. [2]

    Beyondmimic: From motion tracking to versatile humanoid control via guided diffusion,

    Q. Liao, T. E. Truong, X. Huang, Y . Gao, G. Tevet, K. Sreenath, and C. K. Liu, “Beyondmimic: From motion tracking to versatile humanoid control via guided diffusion,”arXiv preprint arXiv:2508.08241, 2025. [Online]. Available: https://github.com/HybridRobotics/whole_body_ tracking

    Pith/arXiv arXiv 2025

  3. [3]

    Retargeting matters: General motion retargeting for humanoid motion tracking,

    J. P. Araujo, Y . Ze, P. Xu, J. Wu, and C. K. Liu, “Retargeting matters: General motion retargeting for humanoid motion tracking,” arXiv preprint arXiv:2510.02252, 2025

    arXiv 2025

  4. [4]

    T. K. Uchida and S. L. Delp,Biomechanics of movement: the science of sports, robotics, and rehabilitation. Mit Press, 2021

    2021

  5. [5]

    Deepmimic: Example-guided deep reinforcement learning of physics-based char- acter skills,

    X. B. Peng, P. Abbeel, S. Levine, and M. Van de Panne, “Deepmimic: Example-guided deep reinforcement learning of physics-based char- acter skills,”ACM Transactions On Graphics (TOG), vol. 37, no. 4, pp. 1–14, 2018

    2018

  6. [6]

    Learning agile robotic locomotion skills by imitating animals,

    X. B. Peng, E. Coumans, T. Zhang, T.-W. E. Lee, J. Tan, and S. Levine, “Learning agile robotic locomotion skills by imitating animals,” in Robotics: Science and Systems, 07 2020

    2020

  7. [7]

    Learning agile skills via adversarial imitation of rough partial demonstrations,

    C. Li, M. Vlastelica, S. Blaes, J. Frey, F. Grimminger, and G. Mar- tius, “Learning agile skills via adversarial imitation of rough partial demonstrations,” inConference on Robot Learning. PMLR, 2023, pp. 342–352

    2023

  8. [8]

    Advanced skills through multiple adversarial motion priors in reinforcement learning,

    E. V ollenweider, M. Bjelonic, V . Klemm, N. Rudin, J. Lee, and M. Hutter, “Advanced skills through multiple adversarial motion priors in reinforcement learning,”arXiv preprint arXiv:2203.14912, 2022

    Pith/arXiv arXiv 2022

  9. [9]

    Asap: Aligning simulation and real-world physics for learning agile humanoid whole-body skills,

    T. He, J. Gao, W. Xiao, Y . Zhang, Z. Wang, J. Wang, Z. Luo, G. He, N. Sobanbab, C. Panet al., “Asap: Aligning simulation and real-world physics for learning agile humanoid whole-body skills,”arXiv preprint arXiv:2502.01143, 2025

    Pith/arXiv arXiv 2025

  10. [10]

    Kungfubot: Physics-based humanoid whole-body control for learning highly-dynamic skills,

    W. Xie, J. Han, J. Zheng, H. Li, X. Liu, J. Shi, W. Zhang, C. Bai, and X. Li, “Kungfubot: Physics-based humanoid whole-body control for learning highly-dynamic skills,”arXiv preprint arXiv:2506.12851, 2025

    Pith/arXiv arXiv 2025

  11. [11]

    Hub: Learning extreme humanoid balance,

    T. Zhang, B. Zheng, R. Nai, Y . Hu, Y .-J. Wang, G. Chen, F. Lin, J. Li, C. Hong, K. Sreenathet al., “Hub: Learning extreme humanoid balance,”arXiv preprint arXiv:2505.07294, 2025

    Pith/arXiv arXiv 2025

  12. [12]

    Omnih2o: Universal and dexterous human- to-humanoid whole-body teleoperation and learning,

    T. He, Z. Luo, X. He, W. Xiao, C. Zhang, W. Zhang, K. Kitani, C. Liu, and G. Shi, “Omnih2o: Universal and dexterous human- to-humanoid whole-body teleoperation and learning,”arXiv preprint arXiv:2406.08858, 2024

    Pith/arXiv arXiv 2024

  13. [13]

    Unitracker: Learning universal whole-body motion tracker for humanoid robots,

    K. Yin, W. Zeng, K. Fan, M. Dai, Z. Wang, Q. Zhang, Z. Tian, J. Wang, J. Pang, and W. Zhang, “Unitracker: Learning universal whole-body motion tracker for humanoid robots,”arXiv preprint arXiv:2507.07356, 2025

    arXiv 2025

  14. [14]

    Twist: Teleoperated whole-body imitation system,

    Y . Ze, Z. Chen, J. P. Araújo, Z.-a. Cao, X. B. Peng, J. Wu, and C. K. Liu, “Twist: Teleoperated whole-body imitation system,”arXiv preprint arXiv:2505.02833, 2025

    Pith/arXiv arXiv 2025

  15. [15]

    Physically based motion transformation,

    Z. Popovi ´c and A. Witkin, “Physically based motion transformation,” inProceedings of the 26th annual conference on Computer graphics and interactive techniques, 1999, pp. 11–20

    1999

  16. [16]

    A physically-based motion retargeting filter,

    S. Tak and H.-S. Ko, “A physically-based motion retargeting filter,” ACM Transactions on Graphics (ToG), vol. 24, no. 1, pp. 98–117, 2005

    2005

  17. [17]

    Ex- pressive whole-body control for humanoid robots,

    X. Cheng, Y . Ji, J. Chen, R. Yang, G. Yang, and X. Wang, “Ex- pressive whole-body control for humanoid robots,”arXiv preprint arXiv:2402.16796, 2024

    Pith/arXiv arXiv 2024

  18. [18]

    Humanplus: Humanoid shadowing and imitation from humans,

    Z. Fu, Q. Zhao, Q. Wu, G. Wetzstein, and C. Finn, “Humanplus: Humanoid shadowing and imitation from humans,”arXiv preprint arXiv:2406.10454, 2024

    Pith/arXiv arXiv 2024

  19. [19]

    Perpetual humanoid control for real-time simulated avatars,

    Z. Luo, J. Cao, A. W. Winkler, K. Kitani, and W. Xu, “Perpetual humanoid control for real-time simulated avatars,” inInternational Conference on Computer Vision (ICCV), 2023

    2023

  20. [20]

    Hover: Versatile neural whole-body controller for humanoid robots,

    T. He, W. Xiao, T. Lin, Z. Luo, Z. Xu, Z. Jiang, C. Liu, G. Shi, X. Wang, L. Fan, and Y . Zhu, “Hover: Versatile neural whole-body controller for humanoid robots,”arXiv preprint arXiv:2410.21229, 2024

    Pith/arXiv arXiv 2024

  21. [21]

    Models, feedback control, and open problems of 3d bipedal robotic walking,

    J. W. Grizzle, C. Chevallereau, R. W. Sinnet, and A. D. Ames, “Models, feedback control, and open problems of 3d bipedal robotic walking,”Automatica, vol. 50, no. 8, pp. 1955–1988, 2014

    1955

  22. [22]

    Algorithmic foundations of realizing multi-contact locomotion on the humanoid robot durus,

    J. P. Reher, A. Hereid, S. Kolathaya, C. M. Hubicki, and A. D. Ames, “Algorithmic foundations of realizing multi-contact locomotion on the humanoid robot durus,” inAlgorithmic F oundations of Robotics XII: Proceedings of the Twelfth Workshop on the Algorithmic F oundations of Robotics. Springer, 2020, pp. 400–415

    2020

  23. [23]

    Spatio-temporal motion retargeting for quadruped robots,

    T. Yoon, D. Kang, S. Kim, J. Cheng, M. Ahn, S. Coros, and S. Choi, “Spatio-temporal motion retargeting for quadruped robots,”IEEE Transactions on Robotics, 2025

    2025

  24. [24]

    Dynaretarget: Dynamically-feasible retargeting using sampling-based trajectory optimization,

    V . Dhedin, I. Taouil, S. Omar, D. Yu, K. Tao, A. Dai, and M. Khadiv, “Dynaretarget: Dynamically-feasible retargeting using sampling-based trajectory optimization,”arXiv preprint arXiv:2602.06827, 2026

    Pith/arXiv arXiv 2026

  25. [25]

    Mujoco: A physics engine for model-based control,

    E. Todorov, T. Erez, and Y . Tassa, “Mujoco: A physics engine for model-based control,” in2012 IEEE/RSJ international conference on intelligent robots and systems. IEEE, 2012, pp. 5026–5033

    2012

  26. [26]

    A compre- hensive, open-source dataset of lower limb biomechanics in multiple conditions of stairs, ramps, and level-ground ambulation and transi- tions,

    J. Camargo, A. Ramanathan, W. Flanagan, and A. Young, “A compre- hensive, open-source dataset of lower limb biomechanics in multiple conditions of stairs, ramps, and level-ground ambulation and transi- tions,”Journal of biomechanics, vol. 119, p. 110320, 2021

    2021

  27. [27]

    Motionpro: exploring the role of pressure in human mocap and beyond,

    S. Ren, Y . Lu, J. Huang, J. Zhao, H. Zhang, T. Yu, Q. Shen, and X. Cao, “Motionpro: exploring the role of pressure in human mocap and beyond,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 27 760–27 770

    2025

  28. [28]

    Groundlink: A dataset unifying human body movement and ground reaction dynamics,

    X. Han, B. Senderling, S. To, D. Kumar, E. Whiting, and J. Saito, “Groundlink: A dataset unifying human body movement and ground reaction dynamics,” inSIGGRAPH Asia 2023 Conference Papers, 2023, pp. 1–10

    2023

  29. [29]

    Physics-informed ground reaction dynamics from human motion capture,

    C. Le, H.-P. Le, D. Le, M.-T. Duong, V .-B. Nguyen, and M.-H. Le, “Physics-informed ground reaction dynamics from human motion capture,” in2025 17th International Conference on Human System Interaction (HSI). IEEE, 2025, pp. 1–6

    2025

  30. [30]

    Softmimic: Learn- ing compliant whole-body control from examples,

    G. B. Margolis, M. Wang, N. Fey, and P. Agrawal, “Softmimic: Learn- ing compliant whole-body control from examples,”arXiv preprint arXiv:2510.17792, 2025

    arXiv 2025

  31. [31]

    Opensim: open-source software to create and analyze dynamic simulations of movement,

    S. L. Delp, F. C. Anderson, A. S. Arnold, P. Loan, A. Habib, C. T. John, E. Guendelman, and D. G. Thelen, “Opensim: open-source software to create and analyze dynamic simulations of movement,” IEEE transactions on biomedical engineering, vol. 54, no. 11, pp. 1940–1950, 2007

    1940

  32. [32]

    CasADi – A software framework for nonlinear optimization and optimal control,

    J. A. E. Andersson, J. Gillis, G. Horn, J. B. Rawlings, and M. Diehl, “CasADi – A software framework for nonlinear optimization and optimal control,”Mathematical Programming Computation, vol. 11, no. 1, pp. 1–36, 2019

    2019

  33. [33]

    The pinocchio c++ library – a fast and flexible implementation of rigid body dynamics algorithms and their analytical derivatives,

    J. Carpentier, G. Saurel, G. Buondonno, J. Mirabel, F. Lamiraux, O. Stasse, and N. Mansard, “The pinocchio c++ library – a fast and flexible implementation of rigid body dynamics algorithms and their analytical derivatives,” inIEEE International Symposium on System Integrations (SII), 2019

    2019

  34. [34]

    Pressurevision++: Estimating fingertip pressure from diverse rgb images,

    P. Grady, J. A. Collins, C. Tang, C. D. Twigg, K. Aneja, J. Hays, and C. C. Kemp, “Pressurevision++: Estimating fingertip pressure from diverse rgb images,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), January 2024, pp. 8698–8708

    2024

This paper was first reviewed by deepseek-v4-flash on August 3, 2026.