pith. sign in

arxiv: 2606.12728 · v1 · pith:5YYLB5ESnew · submitted 2026-06-10 · 💻 cs.RO · cs.CV· cs.LG

EquiDexFlow: Contact-Grounded SE(3)-Equivariant Dexterous Grasp Generative Flows

Clinton Enwerem , John S. Baras , Calin Belta This is my paper

Pith reviewed 2026-06-27 09:18 UTC · model grok-4.3

classification 💻 cs.RO cs.CVcs.LG
keywords dexterous graspingSE(3) equivarianceflow matchingcontact forcesrobotic manipulationgrasp generationforce closureAllegro Hand
0
0 comments X

The pith

EquiDexFlow generates dexterous grasps that satisfy surface contact and Coulomb friction by architectural projection rather than post-verification.

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

The paper introduces EquiDexFlow, an SE(3)-equivariant flow-matching model that takes an object point cloud and outputs wrist pose, joint angles, fingertip contacts, normals, and forces together. Contacts are projected onto the object surface and forces into the friction cone inside the network, so these constraints hold without dedicated loss terms. The architecture is proven equivariant end-to-end and the property is checked empirically across 200 rotations with wrist error below 0.04 degrees and zero joint deviation. Trained on 8100 force-closure grasps from 81 objects for a 16-DoF Allegro Hand, the model records zero friction violations and the lowest wrench residual among tested variants. Retargeted outputs to a LEAP Hand succeed in open-loop physical pick-and-hold trials on all six test objects.

Core claim

EquiDexFlow is an SE(3)-equivariant flow-matching model that jointly predicts wrist pose, joint angles, fingertip contacts, surface normals and contact forces from an object point cloud for the Allegro Hand; built-in projections place contacts on the surface and forces inside the Coulomb cone so that placement and friction compliance are satisfied without loss penalties, yielding zero friction violations, the best composite score and the lowest wrench residual on the training distribution while preserving exact equivariance.

What carries the argument

SE(3)-equivariant flow-matching network whose layers project contacts onto the object surface and forces into the Coulomb friction cone by construction.

If this is right

  • The generated grasps exhibit zero friction violations and the lowest wrench residual among ablation variants.
  • End-to-end SE(3) equivariance holds with wrist residuals below 0.04 degrees and exactly zero joint deviation under 200 rotations.
  • Retargeted contacts to the LEAP Hand keep every joint at least 5 percent inside its actuator limits while preserving wrench balance.
  • Open-loop physical pick-and-hold succeeds on every test object, including asymmetric objects at both canonical and 120-degree rotated poses.

Where Pith is reading between the lines

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

  • The built-in projections may allow grasp models to be trained with fewer explicit stability penalties than current practice.
  • Exact equivariance could reduce the number of object orientations needed during data collection.
  • The same projection mechanism might extend to other constraint types such as collision avoidance or joint limits.

Load-bearing premise

The 8100 force-closure grasps across 81 objects plus the architectural projections are enough to produce stable grasps on unseen objects and under real dynamics without further verification.

What would settle it

A set of grasps generated for objects outside the 81 training objects that produce measurable friction violations or drop during physical pick-and-hold execution would falsify the claim.

Figures

Figures reproduced from arXiv: 2606.12728 by Calin Belta, Clinton Enwerem, John S. Baras.

Figure 1
Figure 1. Figure 1: SE(3)-Equivariant Dexterous Grasps Generated by E [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: EQUIDEXFLOW Architecture. A VN-DGCNN encoder produces equivariant features zO that drive five heads: an SE(3) wrist-pose flow and joint, contact, normal, and force decoders. The normal decoder replaces centroid estimates that shrink the wrench-balance feasible set, and a cone-projection layer maps each force into the friction cone by construction. Wrench, friction, and collision losses supervise the contac… view at source ↗
Figure 3
Figure 3. Figure 3: Contact Geometry and Friction Cone Projection. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Wrist Refinement. Pre-contact IK (left): the hand at the raw decoded wrist pose and joint configu￾ration qˆh hovers above the YCB tomato soup can, and the predicted contacts Cˆ are not realized by the kine￾matic fingertips gi(Tw, qh). Contact IK (right): mini￾mizing (11) seats each fingertip on the surface, driving gi(Tw, qh) → Cˆi. The flow-decoded wrist typically places the fingertips a few cen￾timeters … view at source ↗
Figure 5
Figure 5. Figure 5: Equivariance and Grasp Quality. (a) Grasps co-rotate with the input under Rz rotation (wrist residual <0.04◦ , max ∆qh=0, [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: End-to-End SE(3) Equivariance of EQUIDEXFLOW on the 16-DoF Allegro Hand. Six objects (left to right, then top to bottom in triples: the YCB pudding box, EGAD! G6, the box primitive, the YCB mustard bottle, EGAD! E4, and EGAD! G5) across three input rotations (columns: 0 ◦ , 120◦ , 240◦ about the vertical axis). The wrist pose and finger configuration co-rotate with the object. The wrist residuals stay belo… view at source ↗
Figure 7
Figure 7. Figure 7: Validation Curves (FULL Model). Joint-angle NLL (left, →0.18 nats), SE(3) flow loss (center, →0.9), and total loss (right, ∼11→7.5) with the wrench-balance loss (→1). Loss components in [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Hardware Platform and Equivariant Execution. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Allegro Hand Grasps Generated by EQUIDEXFLOW. Sixteen representative test-set objects spanning EGAD!, YCB, and the cylinder primitive (full inventory in Section E). Each grasp is decoded from the object point cloud by the FULL model and seated on the surface by the wrist refinement of Section III.B. E. DEXTEROUS DATASET PROVENANCE We extend the dataset summary of Section IV with the full per-record schema.… view at source ↗
Figure 10
Figure 10. Figure 10: LEAP Grasp Gallery with Per-Grasp Metrics. [PITH_FULL_IMAGE:figures/full_fig_p019_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Pedestal Registration Jig. Object footprints on the 150 mm-diameter cylindrical pedestal at the canonical pose (left, 0 ◦ ) and after a 120◦ rotation about the vertical axis (right). The four outlines co-rotate rigidly between panels, and B/C/M/P denote the box primitive, cube, mustard bottle, and potted meat can. Printed at full scale and affixed to the pedestal visible in [PITH_FULL_IMAGE:figures/full_… view at source ↗
read the original abstract

Most learned dexterous grasp generators relegate contact forces to a downstream verification step, so a kinematically-plausible pose can still violate the conditions for a stable physical grasp. We address this with EquiDexFlow, an SE(3)-equivariant flow-matching model that jointly predicts wrist pose, joint angles, fingertip contacts, surface normals, and contact forces from an object point cloud. Our architecture projects contacts onto the object surface and forces into the Coulomb friction cone by construction, so placement and friction compliance hold without loss penalties. We prove end-to-end SE(3) equivariance and verify it empirically over 200 rotations, with wrist residuals below $0.04^\circ$ and exactly zero joint deviation. Trained on 8,100 force-closure grasps across 81 objects for the 16-DoF Allegro Hand, our model achieves zero friction violations, the best composite score, and the lowest wrench residual among all ablation variants. We retarget decoded fingertip contacts to a 16-DoF LEAP Hand via per-finger inverse kinematics, and our hardware-feasible refinement places every joint at least 5% inside its actuator envelope while preserving wrench balance. On the physical robot, retargeted EquiDexFlow-decoded grasps complete open-loop pick-and-hold trials on all six test objects, with every asymmetric object succeeding at both the canonical pose and a $120^\circ$ co-rotation. Videos, code, and checkpoints are available at https://equidexflow.github.io.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The manuscript presents EquiDexFlow, an SE(3)-equivariant flow-matching model for dexterous grasp generation. It jointly predicts wrist pose, joint angles, fingertip contacts, surface normals, and contact forces from an object point cloud for the 16-DoF Allegro Hand. The architecture incorporates projections to enforce contact placement on the object surface and forces within the Coulomb friction cone by construction. The paper provides a proof of end-to-end SE(3) equivariance, empirically verifies it over 200 rotations, and reports zero friction violations, superior composite scores, and successful open-loop physical robot trials on six test objects, including rotated asymmetric objects, after retargeting to the LEAP Hand.

Significance. If the central claims hold, this work offers a significant contribution to the field of robotic grasping by integrating physical constraints directly into the generative process through architectural design rather than loss penalties. The proven and verified equivariance, combined with hardware validation, suggests improved generalization and reliability for dexterous manipulation tasks. The release of code, checkpoints, and videos enhances the potential impact and reproducibility.

minor comments (2)
  1. [Abstract] Abstract: the composite score referenced as 'best' is not defined in the abstract; a brief parenthetical or reference to the relevant table/equation would improve clarity for readers.
  2. The source and curation criteria for the 8,100 force-closure grasps (e.g., whether from a public dataset and any exclusion rules) should be stated explicitly in the methods section to support the generalization claims.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive evaluation of EquiDexFlow, the verification of our central claims on equivariance and constraint enforcement, and the recommendation for minor revision. No major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper's central claims rest on an architectural design that projects contacts and forces by construction (enforcing surface placement and Coulomb-cone compliance without loss terms), a mathematical proof of end-to-end SE(3) equivariance, and empirical verification on an external dataset of 8,100 force-closure grasps plus physical robot trials. None of these steps reduce the reported metrics (zero friction violations, lowest wrench residual) to quantities defined by fitted parameters from the same data or to self-citations whose content is unverified. The training data and hardware evaluation provide independent grounding, and the equivariance proof is presented as a self-contained derivation rather than an imported uniqueness result. No self-definitional, fitted-input, or ansatz-smuggling patterns appear in the derivation chain.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available, so free parameters, axioms, and invented entities cannot be exhaustively audited from the text. The approach relies on standard assumptions of flow-matching models and SE(3) group representations from prior literature; no new entities are postulated.

pith-pipeline@v0.9.1-grok · 5828 in / 1418 out tokens · 22159 ms · 2026-06-27T09:18:52.855115+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

42 extracted references · 7 canonical work pages · 4 internal anchors

  1. [1]

    EGAD! An Evolved Grasping Analysis Dataset for Diversity and Reproducibility in Robotic Manipulation.IEEE Robotics and Automation Letters, 5(3):4368–4375, 2020

    Douglas Morrison, Peter Corke, and Jurgen Leitner. EGAD! An Evolved Grasping Analysis Dataset for Diversity and Reproducibility in Robotic Manipulation.IEEE Robotics and Automation Letters, 5(3):4368–4375, 2020

    2020

  2. [2]

    Byeongdo Lim, Jongmin Kim, Jihwan Kim, Yonghyeon Lee, and Frank C. Park. EquiGraspFlow: SE(3)- Equivariant 6-DoF Grasp Pose Generative Flows. InProceedings of The 8th Conference on Robot Learning, volume 270 ofProceedings of Machine Learning Research, pages 5067–5086, 2025. 12 C. ENWEREM ET AL. - EQUIDEXFLOW: SE(3)-EQUIVARIANTDEXTEROUSGRASPGENERATIVEFLOWS

    2025

  3. [3]

    Congyue Deng, Or Litany, Yueqi Duan, Adrien Poulenard, Andrea Tagliasacchi, and Leonidas J. Guibas. Vector Neurons: A General Framework for SO(3)-Equivariant Networks. InProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 12200–12209, 2021

    2021

  4. [4]

    Lipman, R

    Y . Lipman, R. T. Chen, H. Ben-Hamu, M. Nickel, and M. Le. Flow Matching for Generative Modeling. InThe Eleventh International Conference on Learning Representations, 2023

    2023

  5. [5]

    Munthe-Kaas

    H. Munthe-Kaas. Runge-Kutta Methods on Lie Groups.BIT Numerical Mathematics, 38:92–111, 1998

    1998

  6. [6]

    Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Ermon, and B. Poole. Score-Based Generative Modeling through Stochastic Differential Equations. InThe Ninth International Conference on Learning Representations, 2021

    2021

  7. [7]

    Li, Preston Culbertson, Joel W

    Albert H. Li, Preston Culbertson, Joel W. Burdick, and Aaron D. Ames. FRoGGeR: Fast Robust Grasp Gen- eration via the Min-Weight Metric. InIEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023

    2023

  8. [8]

    Tengyu Liu, Zeyu Liu, Ziyuan Jiao, Yixin Zhu, and Song-Chun Zhu. Synthesizing Diverse and Physically Stable Grasps With Arbitrary Hand Structures Using Differentiable Force Closure Estimator.IEEE Robotics and Automation Letters, 7(1):470–477, 2022

    2022

  9. [9]

    DexGraspNet: A Large-Scale Robotic Dexterous Grasp Dataset for General Objects Based on Simulation

    Ruicheng Wang, Jialiang Zhang, Jiayi Chen, Yinzhen Xu, Puhao Li, Tengyu Liu, and He Wang. DexGraspNet: A Large-Scale Robotic Dexterous Grasp Dataset for General Objects Based on Simulation. InProceedings of the IEEE International Conference on Robotics and Automation, pages 11359–11366, 2023

    2023

  10. [10]

    DexGraspNet 2.0: Learning Generative Dexterous Grasping in Large-Scale Synthetic Cluttered Scenes

    Jialiang Zhang, Haoran Liu, Danshi Li, Xinqiang Yu, Haoran Geng, Yufei Ding, Jiayi Chen, and He Wang. DexGraspNet 2.0: Learning Generative Dexterous Grasping in Large-Scale Synthetic Cluttered Scenes. InPro- ceedings of The 8th Conference on Robot Learning, volume 270 ofProceedings of Machine Learning Research, pages 5106–5133, 2025

    2025

  11. [11]

    GenDexGrasp: Generalizable Dexterous Grasping

    Puhao Li, Tengyu Liu, Yuyang Li, Yiran Geng, Yixin Zhu, Yaodong Yang, and Siyuan Huang. GenDexGrasp: Generalizable Dexterous Grasping. In2023 IEEE International Conference on Robotics and Automation (ICRA), pages 8068–8074, 2023

    2023

  12. [12]

    GrainGrasp: Dexterous Grasp Generation with Fine-Grained Contact Guidance

    Fuqiang Zhao, Dzmitry Tsetserukou, and Qian Liu. GrainGrasp: Dexterous Grasp Generation with Fine-Grained Contact Guidance. In2024 IEEE International Conference on Robotics and Automation (ICRA), pages 6470– 6476, 2024

    2024

  13. [13]

    Mousavian, C

    A. Mousavian, C. Eppner, and D. Fox. 6-DoF GraspNet: Variational Grasp Generation for Object Manipulation. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 2901–2910, 2019

    2019

  14. [14]

    Urain, N

    J. Urain, N. Funk, J. Peters, and G. Chalvatzaki. SE(3)-DiffusionFields: Learning Smooth Cost Functions for Joint Grasp and Motion Optimization through Diffusion. In2023 IEEE International Conference on Robotics and Automation (ICRA), pages 5923–5930, 2023

    2023

  15. [15]

    Z. Weng, H. Lu, J. Lundell, and D. Kragic. CapGrasp: AnR 3×SO(2)-Equivariant Continuous Approach- Constrained Generative Grasp Sampler.IEEE Robotics and Automation Letters, 2024

    2024

  16. [16]

    DexDiffuser: Generating Dexterous Grasps With Diffusion Models.IEEE Robotics and Automation Letters, 9(12):11834–11840, 2024

    Zehang Weng, Haofei Lu, Danica Kragic, and Jens Lundell. DexDiffuser: Generating Dexterous Grasps With Diffusion Models.IEEE Robotics and Automation Letters, 9(12):11834–11840, 2024

    2024

  17. [17]

    FFHFlow: A Flow-based Variational Approach for Learning Diverse Dexterous Grasps with Shape-Aware Introspection.Transactions on Machine Learning Research, 2025

    Qian Feng, Jianxiang Feng, Zhaopeng Chen, Rudolph Triebel, and Alois Knoll. FFHFlow: A Flow-based Variational Approach for Learning Diverse Dexterous Grasps with Shape-Aware Introspection.Transactions on Machine Learning Research, 2025

    2025

  18. [18]

    UniDexGrasp: Universal Robotic Dexterous Grasp- ing via Learning Diverse Proposal Generation and Goal-Conditioned Policy

    Yinzhen Xu, Weikang Wan, Jialiang Zhang, Haoran Liu, Zikang Shan, Hao Shen, Ruicheng Wang, Haoran Geng, Yijia Weng, Jiayi Chen, Tengyu Liu, Li Yi, and He Wang. UniDexGrasp: Universal Robotic Dexterous Grasp- ing via Learning Diverse Proposal Generation and Goal-Conditioned Policy. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec...

    2023

  19. [19]

    X. Zhu, D. Wang, O. Biza, G. Su, R. Walters, and R. Platt. Sample Efficient Grasp Learning Using Equivariant Models. InRobotics: Science and Systems, 2022

    2022

  20. [20]

    Huang, D

    H. Huang, D. Wang, X. Zhu, R. Walters, and R. Platt. Edge Grasp Network: A Graph-Based SE(3)-Invariant Approach to Grasp Detection. In2023 IEEE International Conference on Robotics and Automation (ICRA), pages 3882–3888, 2023

    2023

  21. [21]

    Köhler, L

    J. Köhler, L. Klein, and F. Noé. Equivariant Flows: Exact Likelihood Generative Learning for Symmetric Densities. InInternational Conference on Machine Learning, pages 5361–5370. PMLR, 2020

    2020

  22. [22]

    Katsman, A

    I. Katsman, A. Lou, D. Lim, Q. Jiang, S. N. Lim, and C. M. De Sa. Equivariant Manifold Flows.Advances in Neural Information Processing Systems, 34:10600–10612, 2021. 13 C. ENWEREM ET AL. - EQUIDEXFLOW: SE(3)-EQUIVARIANTDEXTEROUSGRASPGENERATIVEFLOWS

    2021

  23. [23]

    K. Chen, X. Chen, Z. Yu, M. Zhu, and H. Yang. EquiDiff: A Conditional Equivariant Diffusion Model for Trajectory Prediction. In2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), pages 746–751, 2023

    2023

  24. [24]

    Rozenberg and D

    E. Rozenberg and D. Freedman. Semi-Equivariant Conditional Normalizing Flows, with Applications to Target- Aware Molecule Generation.Machine Learning: Science and Technology, 4(3):035037, 2023

    2023

  25. [25]

    EFF-Grasp: Energy-Field Flow Matching for Physics-Aware Dexterous Grasp Generation, 2026

    Yukun Zhao, Zichen Zhong, Yongshun Gong, Yilong Yin, and Haoliang Sun. EFF-Grasp: Energy-Field Flow Matching for Physics-Aware Dexterous Grasp Generation, 2026. Preprint, arXiv:2603.16151

    work page arXiv 2026

  26. [26]

    GraspSense: Physically Grounded Grasp and Grip Planning for a Dexterous Robotic Hand via Language-Guided Perception and Force Maps

    Elizaveta Semenyakina, Ivan Snegirev, Mariya Lezina, Miguel Altamirano Cabrera, Safina Gulyamova, and Dzmitry Tsetserukou. GraspSense: Physically Grounded Grasp and Grip Planning for a Dexterous Robotic Hand via Language-Guided Perception and Force Maps, 2026. Preprint, arXiv:2604.05697

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  27. [27]

    Adversarial Game-Theoretic Algorithm for Dexterous Grasp Synthesis, 2025

    Yu Chen, Botao He, Yuemin Mao, Arthur Jakobsson, Jeffrey Ke, Yiannis Aloimonos, Guanya Shi, Howie Choset, Jiayuan Mao, and Jeffrey Ichnowski. Adversarial Game-Theoretic Algorithm for Dexterous Grasp Synthesis, 2025

    2025

  28. [28]

    Risk-Constrained Belief-Space Optimization for Safe Control under Latent Uncertainty

    Clinton Enwerem, John S. Baras, and Calin Belta. Risk-Constrained Belief-Space Optimization for Safe Control under Latent Uncertainty, 2026. arXiv:2604.03868

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  29. [29]

    Variational Neural Belief Parameterizations for Robust Dexterous Grasping under Multimodal Uncertainty

    Clinton Enwerem, Shreya Kalyanaraman, John S. Baras, and Calin Belta. Variational Neural Parameterizations for Robust Dexterous Grasping under Multimodal Uncertainty, 2026. arXiv:2604.25897

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  30. [30]

    A micro Lie theory for state estimation in robotics,

    Joan Solà, Jeremie Deray, and Dinesh Atchuthan. A Micro Lie Theory for State Estimation in Robotics, 2021. Preprint, arXiv:1812.01537

    work page arXiv 2021

  31. [31]

    Murray, Zexiang Li, and S

    Richard M. Murray, Zexiang Li, and S. Shankar Sastry.A Mathematical Introduction to Robotic Manipulation. CRC Press, 1st edition, 2017

    2017

  32. [32]

    R. T. Q. Chen and Y . Lipman. Flow Matching on General Geometries. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  33. [33]

    R. T. Chen, Y . Rubanova, J. Bettencourt, and D. K. Duvenaud. Neural Ordinary Differential Equations.Advances in Neural Information Processing Systems, 31:6572–6583, 2018

    2018

  34. [34]

    Mathieu and M

    E. Mathieu and M. Nickel. Riemannian Continuous Normalizing Flows.Advances in Neural Information Pro- cessing Systems, 33:2503–2515, 2020

    2020

  35. [35]

    Classifier-Free Diffusion Guidance

    Jonathan Ho and Tim Salimans. Classifier-Free Diffusion Guidance, 2022. Preprint, arXiv:2207.12598

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  36. [36]

    Density Estimation Using Real NVP

    Laurent Dinh, Jascha Sohl-Dickstein, and Samy Bengio. Density Estimation Using Real NVP. InInternational Conference on Learning Representations, 2017

    2017

  37. [37]

    Drake: Model-Based Design and Verification for Robotics, 2019

    Russ Tedrake and the Drake Development Team. Drake: Model-Based Design and Verification for Robotics, 2019

    2019

  38. [38]

    Berk Calli, Aaron Walsman, Arjun Singh, Siddhartha Srinivasa, Pieter Abbeel, and Aaron M. Dollar. The YCB Object and Model Set: Towards Common Benchmarks for Manipulation Research. In2015 International Conference on Advanced Robotics (ICAR), pages 510–517, 2015

    2015

  39. [39]

    D. P. Kingma and J. Ba. Adam: A Method for Stochastic Optimization. InThe Third International Conference on Learning Representations, 2015

    2015

  40. [40]

    LEAP Hand: Low-Cost, Efficient, and Anthropomorphic Hand for Robot Learning

    Kenneth Shaw, Ananye Agarwal, and Deepak Pathak. LEAP Hand: Low-Cost, Efficient, and Anthropomorphic Hand for Robot Learning. InRobotics: Science and Systems XIX, 2023

    2023

  41. [41]

    Umit Rusen Aktas, Chao Zhao, Marek Kopicki, Ales Leonardis, and Jeremy L. Wyatt. Deep Dexterous Grasping of Novel Objects from a Single View, 2019

    2019

  42. [42]

    GAGrasp: Geometric Algebra Diffusion for Dexterous Grasping

    Tao Zhong and Christine Allen-Blanchette. GAGrasp: Geometric Algebra Diffusion for Dexterous Grasping. In IEEE International Conference on Robotics and Automation (ICRA), 2025. arXiv:2503.04123. 14 C. ENWEREM ET AL. - EQUIDEXFLOW: SE(3)-EQUIVARIANTDEXTEROUSGRASPGENERATIVEFLOWS APPENDIX A. ARCHITECTUREDETAILS The VN-MLP is parameterized by four VN-Linear-L...

    work page arXiv 2025