Pith. sign in

REVIEW 4 major objections 6 minor 27 references

Physically Consistent Humanoid Loco-Manipulation using Latent Diffusion Models

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read This paper claims that latent-diffusion-generated images can supply the contact points and robot poses that let whole-body trajectory optimization solve long-horizon humanoid loco-manipulation in simulation.

desk verdict A genuine systems contribution with a load-bearing geometric-transfer step that is under-validated; deserves review but needs sensitivity analysis and more scenarios. read the letter →

arxiv 2504.16843 v1 pith:LPLDQVG5 submitted 2025-04-23 cs.RO cs.GR

Ilyass Taouil , Haizhou Zhao , Angela Dai , Majid Khadiv This is my paper · ORCID
classification cs.ROcs.GR
keywords humanoidloco-manipulationlatentdiffusionmodelstrajectoryoptimizationcontactplanningwhole-bodycontrolimage-guidedretargetinglong-horizonsimulationvalidation
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 show that a text-to-image latent diffusion model can act as an automatic demonstration generator for humanoid robots. Instead of hand-coding contact sequences or collecting 3D interaction datasets, the method generates RGB images of a person performing a task, extracts from those images where the hands and feet touch objects and what the body pose is, and feeds that information into a whole-body trajectory optimizer. The paper argues that this is enough to produce physically consistent, collision-free trajectories for long-horizon tasks in simulation, including tasks that local trajectory optimization cannot solve when given only semantic contact labels. A sympathetic reader cares because it points to a way of planning complicated loco-manipulation without expensive teleoperation or task-specific heuristics.

What carries the argument

The load-bearing mechanism is a three-stage image-to-optimization pipeline. First, a latent diffusion model generates full-body RGB images of a person carrying out each subtask from a text prompt. Second, a contact-transfer stage lifts 2D hand and foot locations from those images into 3D contact points by matching the generated objects to the simulated scene semantically and then refining the matches with point-cloud geometry and iterative closest point; depth comes from a metric-depth estimator and the missing camera intrinsics are set empirically to the image resolution and half-focal-length principal point offsets. Third, a retargeting stage uses inverse kinematics to map the estimated human pose to the robot's degrees of freedom. The extracted contacts and robot configurations enter the trajectory optimization as keyframe costs and warm starts, which is what turns a local whole-body optimizer into a long-horizon planner.

What would settle it

Take a set of generated images paired with a known 3D scene, for example by rendering a simulated scene from a known camera pose and generating an image of the same object, run the contact-transfer stage, and measure the distance from each extracted contact to the true object surface. If the median distance is comparable to the object dimensions rather than a few centimeters, the contacts cannot guide the optimizer, and the central claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that the plan for a long-horizon loco-manipulation task can be read off from latent-diffusion-generated RGB images: contact locations and robot configurations extracted from those images serve as keyframes that guide a whole-body trajectory optimization with centroidal dynamics and full kinematics. In the two simulated scenarios tested (stepping onto a box to retrieve a basket from a shelf, and loading a box onto a trolley and pushing it), the full pipeline produces trajectories with little or no penetration, whereas a baseline that feeds only semantic contact labels either fails to converge or produces large penetrations. The paper also demonstrates through ablations that both the geometry-refined contact transfer and the keyframe information (base pose, foot position, and subtask-like warm start) contribute to the optimizer's success.

Load-bearing premise

The whole pipeline rests on the assumption that the 3D contact points lifted from the generated image reliably land on the real object's surface; if the estimated depth or the assumed camera intrinsics are wrong, the contacts and keyframes are wrong and the optimizer is led into poor local minima.

Editorial extensions

If this is right

  • If the claim is right, a high-level text description and RGB-D images of the objects are enough to plan long-horizon humanoid loco-manipulation, without task-specific heuristics or 3D interaction datasets.
  • The same generated keyframes let whole-body trajectory optimization solve tasks that local trajectory optimization cannot solve from semantic contact labels alone.
  • Geometry-aware refinement makes the contact transfer robust enough that only a minimal set of collision constraints is needed for collision-free motion.
  • Including base pose, foot position, and warm-start keyframes improves both the convergence rate and the penetration scores of the optimizer across different object masses and initial robot yaws.
  • The approach is demonstrated on two scenarios, so the claimed mechanism transfers across different object types and long-horizon task structures.

Reading between the lines

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

  • Editorial inference: the critical transfer point is the assumed camera intrinsics for the generated image; if this empirical setting does not hold outside the tested simulator, the extracted 3D contacts would be systematically misplaced, so real-world deployment would likely need a calibration or distortion-correction step.
  • Editorial inference: the method suggests a general recipe in which generative image models provide rough demonstrations that seed trajectory optimization; the same pattern could apply to other platforms or tasks where demonstrations are hard to obtain, such as quadruped manipulation or dexterous hand tasks.
  • Editorial inference: because the geometry-overlap criterion is object-agnostic, the contact-transfer stage could be reused with any semantic matcher and depth estimator; a natural test is to swap those components and see whether end-to-end success tracks the contact-transfer error.
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 an end-to-end pipeline that uses latent diffusion models (LDMs) to generate RGB images of human-object interaction scenes, then extracts 2D hand/feet contacts and human poses from those images, lifts them to 3D using segmentation, zero-shot metric depth estimation, and assumed camera intrinsics, and retargets them to a Unitree G1 humanoid in MuJoCo. The resulting contact locations and keyframe configurations are fed into a whole-body trajectory optimization (TO) formulation. The method is evaluated on two long-horizon scenarios (basket retrieval with box stepping, and trolley pushing), with comparisons against contact-only and geometry-unaware baselines and ablations on the contact transfer and keyframe components. The central claim is that LDM-derived contacts and configurations make otherwise intractable local TO problems solvable and reduce collision penetration.

Significance. If the claims hold, the paper makes a useful systems contribution by showing a practical way to inject semantic long-horizon knowledge from generative image models into contact-implicit whole-body TO, without task-specific contact heuristics. The pipeline is clearly described, the use of geometric overlap to refine semantic correspondences is well motivated, and the paper includes ablations for both the contact transfer stage and the keyframe utilization. I also see no circularity: the keyframes and contacts are computed before and independently of the TO, and the evaluation metrics are downstream. However, the evaluation is not yet at the standard needed to support the strong claims: only two scenarios are tested, the plots appear to show single runs without error bars or seed counts, and the 3D lifting step for LDM-generated images is unvalidated. These issues are load-bearing and fixable with additional experiments and analysis.

major comments (4)
  1. [Sec. III-A.2] The contact-transfer stage rests on two unvalidated assumptions: that zero-shot metric depth from Metric3D v2 yields correct metric point clouds for LDM-generated RGB images, and that the hand-set pinhole intrinsics (focal lengths equal to the image resolution and principal-point offsets equal to half the focal lengths, chosen by 'empirical trial and error') are adequate. The rigid transform computed by SVD and ICP from these lifted point clouds is used both to map hand and feet detections to 3D contacts and, in Sec. III-A.3, to define the base orientation for retargeting. If the assumed intrinsics or the estimated depth introduce scale, aspect-ratio, or perspective distortion, the geometric-overlap criterion can select a biased rigid transform, corrupting both the contacts and the keyframe configurations. The paper reports no sensitivity analysis for the intrinsics, no comparison against ground-truth depth for the simulated objects, and no direct accuracy metric for the lifted geometry, so the central claim that LDM images reliably transfer contacts and configurations into simulation is currently supported only by downstream penetration plots. Please add ablations over the intrinsics and depth sources, and report quantitative contact/keyframe errors.
  2. [Sec. V] The quantitative claims that the pipeline 'clearly' reduces penetration and improves TO convergence are made from Figs. 5 and 6, which appear to show single runs with no error bars, seed counts, or variance information. The pipeline is stochastic at multiple stages: LDM image generation, VLM/SAM detection, the sampling-based semantic correspondence search, and possibly the 3D pose estimator, and TO is a local solver sensitive to initialization. Without repeated runs, statements about success rate and comparisons between keyframe settings are not statistically supported. Please report the number of seeds or runs, per-run outcomes, and error bars or box plots, and define the success criterion (e.g., TO convergence, collision-free trajectory, or task completion).
  3. [Sec. V-A] The 'naive' baseline that provides contacts obtained directly from the semantic-aware foundation model is not fully specified. It is unclear how those 2D semantic correspondences are converted to 3D contact locations, which depth source and camera intrinsics are used for the simulated and generated images, and whether the same human-pose extraction and retargeting steps are applied. Without this specification, the comparison in Fig. 5 may conflate the contribution of the geometric refinement with differences in 3D lifting or keyframe usage. Please specify the baseline precisely, ideally by using the same 3D lifting code with only the semantic-correspondence refinement removed.
  4. [Sec. V-C] The keyframe ablation is hard to interpret as reported. The success of the full keyframe set is not uniform: several subtitles in Fig. 6 show other settings with lower or equal penetration, and some bars are marked 'Unsolved' without a definition of what counts as unsolved. The discussion attributes the anomalies to 'some details in the trajectory optimization solver' without further analysis. Please report convergence status and success rates per condition, define the unsolved criterion, and analyze the cases where the full keyframe set performs worse rather than attributing them to unspecified solver details.
minor comments (6)
  1. [Fig. 5] The caption lists '(c) S2 w/ collision penalties' and '(d) S2 w/ collision penalties' identically; one of these should presumably read 'w/o collision penalties' to match the text in Sec. V-B.
  2. [Sec. V-C.2] There is a typo in 'A minimal set of collision constraints are are added'; please remove the duplicated 'are'.
  3. [Table I] The notation 'call' in the first row is unclear; it appears to denote the center of mass of the supported mass, but the acronym and the subscript are not defined in the table or the surrounding text.
  4. [Sec. IV-D] The homotopy parameter alpha in Eq. (13) is introduced without explaining how it is initialized, how it is driven to zero, and whether it is a decision variable or a fixed schedule; please clarify.
  5. [Fig. 6] The legend labels 'n f b bf fw bwbfw' are cryptic; please spell out each abbreviation in the caption so the reader can map the bars to the settings described in Sec. V-C.
  6. [References] The reference for the Flux LDM lists the year as '2014, 2024-07-01'; the 2014 date appears to be an error and should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: LDM-derived keyframes are external inputs to trajectory optimization and are evaluated independently in MuJoCo.

full rationale

Walking the derivation chain, I find no step in which the paper defines a quantity in terms of its own output, fits a parameter and then renames that fit a prediction, or imports a load-bearing claim solely from an overlapping-author citation. The LDM-generated images are produced before contact extraction; the 3D lifts use zero-shot depth plus hand-set intrinsics; the rigid transform is solved by SVD/ICP against the simulated object point cloud; keyframes are extracted once and then used as costs and warm starts in the whole-body TO; and the evaluation metric (collision penetration) is computed in MuJoCo, not optimized by the extraction pipeline. The ablation comparisons (geometry-unaware vs. geometry-refined; keyframe variants) compare trajectories generated from different inputs under the same optimizer, so the reported differences are not forced by construction. The only overlapping-author citation ([14], Khadiv) appears in Related Work and is not used to justify the central premise. Concerns about unvalidated assumed camera intrinsics are validity/correctness risks, not circularity, because no result is defined in terms of those assumptions and no predicted quantity is fed back into its own derivation. I therefore assign 0.

Assumptions & free parameters 4 free parameters · 7 assumptions · 0 invented entities

The central claim rests on domain assumptions about the reliability of image generation, depth estimation, semantic correspondence, and the given high-level plan, plus multiple hand-tuned cost and penalty weights. No new physical entities or conserved quantities are introduced; the humanoid and objects are modeled with standard rigid-body/centroidal dynamics.

free parameters (4)
  • Stagewise TO cost weights (Table I) = Multiple values, e.g., CoM-foot 5e2, leg symmetry 3e2, foot clearance 2e3, slack penalty 1e5
    Hand-chosen quadratic weights in Lstage; they shape gait and stance, and are not derived from data or from the LDM output.
  • Keyframe cost weights W_b_kf and W_f_kf = [100, 10] and 5e2
    Set by hand (Sec. IV-C.1); balance keyframe base-pose and foot-position tracking against the stagewise costs.
  • Collision penalty weights = whom=2e2; others given in Sec. V-A
    Hand-tuned penalty weights for collision avoidance, listed in Sec. IV-D and Sec. V-A.
  • Camera intrinsics for generated images = Focal length = LDM image resolution, principal point = half focal length
    Chosen by 'empirical trial and error' (Sec. III-A.2); directly affects the 3D lifting of generated images and therefore the contact positions L.
assumptions (7)
  • domain assumption A high-level plan P (ordered sequence of text prompts) is provided for each task.
    The planner receives P as input (Sec. III-A) and the paper does not generate the plan itself; the plan is assumed to decompose the long-horizon task correctly.
  • domain assumption For placement tasks, the target 3D location and yaw of the object are given.
    Stated in Sec. III-A; the pipeline does not infer the placement target from the scene.
  • domain assumption The appended text prompt ('A scene of person ... dark hair and casual clothes') forces the LDM to generate full-body, correctly interacting scenes.
    Sec. III-A.1 relies on this to make pose estimation and contact extraction possible; no systematic validation of prompt-dependent failure is provided.
  • domain assumption Correct semantic matches between generated and simulated objects will yield good point-cloud overlap; the highest-overlap rigid transform from the sampled semantic pool is the correct contact transfer.
    Core geometric assumption of the sampling-based contact transfer (Sec. III-A.2), used to refine semantic correspondences with ICP and SVD.
  • domain assumption The centroidal dynamics model (Eq. 1) with semi-implicit Euler discretization is an adequate representation of the humanoid's physics for planning.
    The TO formulation (Sec. IV) rests on this model from [27]; no error analysis against full rigid-body dynamics is reported.
  • domain assumption The contact surface is perpendicular to gravity for patch contacts.
    Eq. (2a) assumes ground/table contacts are horizontal; this is valid in the two simulated scenarios but restricts generality.
  • domain assumption Simulation in MuJoCo with the Unitree G1 model is a sufficient validation platform for physical consistency.
    Results are only in simulation; real-world transfer is explicitly listed as future work in the conclusion.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Physically Consistent Humanoid Loco-Manipulation using Latent Diffusion Models." pith.science (2026). https://pith.science/paper/LPLDQVG5

@misc{pith2026250416843,
  author       = {Pith},
  title        = {Pith review of: Physically Consistent Humanoid Loco-Manipulation using Latent Diffusion Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LPLDQVG5}},
  note         = {Machine review of arXiv:2504.16843}
}
read the original abstract

This paper uses the capabilities of latent diffusion models (LDMs) to generate realistic RGB human-object interaction scenes to guide humanoid loco-manipulation planning. To do so, we extract from the generated images both the contact locations and robot configurations that are then used inside a whole-body trajectory optimization (TO) formulation to generate physically consistent trajectories for humanoids. We validate our full pipeline in simulation for different long-horizon loco-manipulation scenarios and perform an extensive analysis of the proposed contact and robot configuration extraction pipeline. Our results show that using the information extracted from LDMs, we can generate physically consistent trajectories that require long-horizon reasoning.

Figures

Figures reproduced from arXiv: 2504.16843 by the authors.

Figure 1
Figure 1. A loco-manipulation task achieved with our approach. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Pipeline overview. the simulated objects. To solve the semantic correspondence problem, we use a semantic-aware foundation model [24] to obtain semantic matches between the images. However, solely relying on the model is not reliable, as the intra￾class variation can be large leading to incorrect mappings that deteriorate the output trajectory (Sec. V-B). Therefore, we propose a sampling-based algorithm that refines… view at source ↗
Figure 3
Figure 3. Contact extraction procedure [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Keyframe extraction procedure. 35D vector describing the robot’s 6D base state and joint angles for each actuated joint of the system. Extracting this configuration is an important step in our proposed pipeline, as this complementary information to the contacts guides …
Figure 5
Figure 5. Figure 5: Collision penetrations comparison between the TO [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Collision ablation study data. Subtitles are (box mass, robot initial yaw). The upward solid triangles denote that [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 13 canonical work pages

  1. [1]

    Humanoid robot locomotion and manipulation step planning,

    K. Bouyarmane and A. Kheddar, “Humanoid robot locomotion and manipulation step planning,” Advanced Robotics, vol. 26, no. 10, pp. 1099–1126, 2012

    work page 2012

  2. [2]

    A multimode teleoperation framework for humanoid loco- manipulation: An application for the icub robot,

    L. Penco, N. Scianca, V . Modugno, L. Lanari, G. Oriolo, and S. Ivaldi, “A multimode teleoperation framework for humanoid loco- manipulation: An application for the icub robot,” IEEE Robotics & Automation Magazine, vol. 26, no. 4, pp. 73–82, 2019

    work page 2019

  3. [3]

    A standardized benchmark for humanoid whole-body manipulation,

    W. Thibault, F. J. A. Chavez, and K. Mombaur, “A standardized benchmark for humanoid whole-body manipulation,” in 2022 IEEE- RAS 21st International Conference on Humanoid Robots (Humanoids). IEEE, 2022, pp. 608–615

    work page 2022

  4. [4]

    Multi-contact mpc for dynamic loco- manipulation on humanoid robots,

    J. Li and Q. Nguyen, “Multi-contact mpc for dynamic loco- manipulation on humanoid robots,” in 2023 American Control Con- ference (ACC). IEEE, 2023, pp. 1215–1220

    work page 2023

  5. [5]

    Motion primitive based random planning for loco-manipulation tasks,

    A. Settimi, D. Caporale, P. Kryczka, M. Ferrati, and L. Pallot- tino, “Motion primitive based random planning for loco-manipulation tasks,” in 2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids). IEEE, 2016, pp. 1059–1066

    work page 2016

  6. [6]

    Humanoid whole-body planning for loco-manipulation tasks,

    P. Ferrari, M. Cognetti, and G. Oriolo, “Humanoid whole-body planning for loco-manipulation tasks,” in 2017 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2017, pp. 4741–4746

    work page 2017

  7. [7]

    Humanoid loco-manipulation planning based on graph search and reachability maps,

    M. Murooka, I. Kumagai, M. Morisawa, F. Kanehiro, and A. Kheddar, “Humanoid loco-manipulation planning based on graph search and reachability maps,” IEEE Robotics and Automation Letters , vol. 6, no. 2, pp. 1840–1847, 2021

    work page 2021

  8. [8]

    Roloma: Robust loco-manipulation for quadruped robots with arms,

    H. Ferrolho, V . Ivan, W. Merkt, I. Havoutis, and S. Vijayakumar, “Roloma: Robust loco-manipulation for quadruped robots with arms,” Autonomous Robots, vol. 47, no. 8, pp. 1463–1481, 2023

    2023

Show all 27 references
  1. [9]

    Versatile multicontact planning and control for legged loco-manipulation,

    J.-P. Sleiman, F. Farshidian, and M. Hutter, “Versatile multicontact planning and control for legged loco-manipulation,” Science Robotics, vol. 8, no. 81, p. eadg5014, 2023. n f b bf fwbwbfw 0 5 Avg. Penetr. [cm] 2.5 kg, 0 rad n f b bf fwbwbfw 0 5 2.5 kg, 0.6 rad n f b bf fwbw...

    2023

  2. [10]

    Differentiable physics and stable modes for tool-use and manipulation planning,

    M. A. Toussaint, K. R. Allen, K. A. Smith, and J. B. Tenenbaum, “Differentiable physics and stable modes for tool-use and manipulation planning,” Robotics: Science and Systems Foundation , 2018

    2018

  3. [11]

    Deep whole-body control: learning a unified policy for manipulation and locomotion,

    Z. Fu, X. Cheng, and D. Pathak, “Deep whole-body control: learning a unified policy for manipulation and locomotion,” in Conference on Robot Learning. PMLR, 2023, pp. 138–149

    2023

  4. [12]

    Learn- ing whole-body manipulation for quadrupedal robot,

    S. Jeon, M. Jung, S. Choi, B. Kim, and J. Hwangbo, “Learn- ing whole-body manipulation for quadrupedal robot,” arXiv preprint arXiv:2308.16820, 2023

    2023 arXiv

  5. [13]

    Learning force control for legged manipulation,

    T. Portela, G. B. Margolis, Y . Ji, and P. Agrawal, “Learning force control for legged manipulation,” arXiv preprint arXiv:2405.01402 , 2024

    2024 arXiv

  6. [14]

    Learning-based legged locomotion: State of the art and future perspectives,

    S. Ha, J. Lee, M. van de Panne, Z. Xie, W. Yu, and M. Khadiv, “Learning-based legged locomotion: State of the art and future perspectives,” The International Journal of Robotics Research , p. 02783649241312698, 2024

    2024

  7. [15]

    Deep imitation learning for humanoid loco-manipulation through human teleoperation,

    M. Seo, S. Han, K. Sim, S. H. Bang, C. Gonzalez, L. Sentis, and Y . Zhu, “Deep imitation learning for humanoid loco-manipulation through human teleoperation,” in 2023 IEEE-RAS 22nd International Conference on Humanoid Robots (Humanoids). IEEE, 2023, pp. 1–8

    2023

  8. [16]

    Generalizable humanoid manipulation with improved 3d diffusion policies,

    Y . Ze, Z. Chen, W. Wang, T. Chen, X. He, Y . Yuan, X. B. Peng, and J. Wu, “Generalizable humanoid manipulation with improved 3d diffusion policies,” arXiv preprint arXiv:2410.10803 , 2024

    2024 arXiv

  9. [17]

    Mobile aloha: Learning bimanual mobile manipulation with low-cost whole-body teleoperation,

    Z. Fu, T. Z. Zhao, and C. Finn, “Mobile aloha: Learning bimanual mobile manipulation with low-cost whole-body teleoperation,” arXiv preprint arXiv:2401.02117, 2024

    2024 arXiv

  10. [18]

    Opt2skill: Imitating dynamically-feasible whole-body trajectories for versatile humanoid loco-manipulation,

    F. Liu, Z. Gu, Y . Cai, Z. Zhou, S. Zhao, H. Jung, S. Ha, Y . Chen, D. Xu, and Y . Zhao, “Opt2skill: Imitating dynamically-feasible whole-body trajectories for versatile humanoid loco-manipulation,” arXiv preprint arXiv:2409.20514, 2024

    2024

  11. [19]

    [Online]

    black forest labs, “Flux,” 2014, 2024-07-01. [Online]. Available: https://blackforestlabs.ai/

    2014

  12. [20]

    Florence-2: Advancing a unified representation for a variety of vision tasks,

    B. Xiao, H. Wu, W. Xu, X. Dai, H. Hu, Y . Lu, M. Zeng, C. Liu, and L. Yuan, “Florence-2: Advancing a unified representation for a variety of vision tasks,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 4818–4829

    2024

  13. [21]

    Sam 2: Segment anything in images and videos,

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. R ¨adle, C. Rolland, L. Gustafson et al., “Sam 2: Segment anything in images and videos,” arXiv preprint arXiv:2408.00714 , 2024

    2024 arXiv

  14. [22]

    Mujoco: A physics engine for model-based control,

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

    2012

  15. [23]

    Metric3d v2: A versatile monocular geometric foundation model for zero-shot metric depth and surface normal estimation,

    M. Hu, W. Yin, C. Zhang, Z. Cai, X. Long, H. Chen, K. Wang, G. Yu, C. Shen, and S. Shen, “Metric3d v2: A versatile monocular geometric foundation model for zero-shot metric depth and surface normal estimation,” arXiv preprint arXiv:2404.15506 , 2024

    2024 arXiv

  16. [24]

    Telling left from right: Identifying geometry-aware semantic correspondence,

    J. Zhang, C. Herrmann, J. Hur, E. Chen, V . Jampani, D. Sun, and M.- H. Yang, “Telling left from right: Identifying geometry-aware semantic correspondence,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024

    2024

  17. [25]

    Wham: Reconstructing world-grounded humans with accurate 3d motion,

    S. Shin, J. Kim, E. Halilaj, and M. J. Black, “Wham: Reconstructing world-grounded humans with accurate 3d motion,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2024, pp. 2070–2080

    2024

  18. [26]

    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

    2024 arXiv

  19. [27]

    Whole-body motion planning with centroidal dynamics and full kinematics,

    H. Dai, A. Valenzuela, and R. Tedrake, “Whole-body motion planning with centroidal dynamics and full kinematics,” in 2014 IEEE-RAS International Conference on Humanoid Robots , 2014, pp. 295–302

    2014

Pith tools

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