DexMachina: Functional Retargeting for Bimanual Dexterous Manipulation
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:OQZNUT6Grecord.jsonopen to challenge →
read the original abstract
We study the problem of functional retargeting: learning dexterous manipulation policies to track object states from human hand-object demonstrations. We focus on long-horizon, bimanual tasks with articulated objects, which is challenging due to large action space, spatiotemporal discontinuities, and embodiment gap between human and robot hands. We propose DexMachina, a novel curriculum-based algorithm: the key idea is to use virtual object controllers with decaying strength: an object is first driven automatically towards its target states, such that the policy can gradually learn to take over under motion and contact guidance. We release a simulation benchmark with a diverse set of tasks and dexterous hands, and show that DexMachina significantly outperforms baseline methods. Our algorithm and benchmark enable a functional comparison for hardware designs, and we present key findings informed by quantitative and qualitative results. With the recent surge in dexterous hand development, we hope this work will provide a useful platform for identifying desirable hardware capabilities and lower the barrier for contributing to future research. Videos and more at https://project-dexmachina.github.io/
This paper has not been read by Pith yet.
Forward citations
Cited by 16 Pith papers
-
Human Universal Grasping
HUG trains a flow-matching model on a new 1M-frame egocentric human grasp dataset to generate retargetable grasps from single RGB-D images, beating baselines by 23-34% on a new 90-object benchmark.
-
EgoEngine: From Egocentric Human Videos to High-Fidelity Dexterous Robot Demonstrations
EgoEngine transforms egocentric human videos into high-fidelity robot data enabling zero-shot visuomotor dexterous policy learning without real-robot demonstrations.
-
DexFuture: Hierarchical Future-State Visuomotor Targeting for Bimanual Dexterous Tool Use
DexFuture reaches 90% of oracle performance on bimanual tool-use tasks at 60 Hz by using a horizon-conditioned transformer to predict future visuomotor targets and a per-link policy to track them.
-
WristMimic: Full-Body Humanoid Control with Wrist-Guided Manipulation
WristMimic achieves comparable or superior object manipulation retargeting by supervising wrist kinematics while letting finger behavior emerge from object and contact dynamics.
-
Support-Constrained RL Enables Real-World Policy Improvement without Real-World Experience
SCORE constrains sim RL to the support of a real-data policy via flow steering, raising average success on eight dexterous tasks from 37.8% to 89.9%.
-
Do as I Do: Dexterous Manipulation Data from Everyday Human Videos
DO AS I DO reconstructs and retargets hand-object interactions from in-the-wild monocular RGB videos to produce dexterous robot manipulation trajectories, outperforming prior methods on ground-truth and online video datasets.
-
Hand-centric Human-to-Robot Trajectory Transfer from Video Demonstrations via Open-World Contact Localization
HOWTransfer recovers 3D hand motion from video, localizes contact intervals via hand-object cues, generates multi-modal grasp hypotheses, and edits trajectories to produce diverse robot-executable motions achieving 86...
-
Video2Sim2Real: Full-Stack Autonomous Dexterous Skill Acquisition from a Single Human Video
Video2Sim2Real turns a single human video into a deployable robot manipulation skill by reconstructing a digital twin, anchoring motions to object-centric simulator configurations, and bridging sim-to-real gaps with i...
-
NeuROK: Generative 4D Neural Object Kinematics
NeuROK learns a data-driven latent kinematic parameterization on a large 4D dataset to generate realistic object deformations by simulating dynamics only in low-dimensional latent space via Lagrangian mechanics.
-
DexTwist: Dexterous Hand Retargeting for Twist Motion via Mixed Reality-based Teleoperation
DexTwist detects tripod pinches, estimates the intended screw axis and twist magnitude, then applies real-time joint refinement to track turning progress while stabilizing the robot's tripod geometry.
-
Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?
Play2Perfect uses task-agnostic RL play pretraining on diverse objects to build reusable manipulation priors, then fine-tunes for assembly, yielding 33x sample efficiency gains and 60% success on 0.5mm-clearance inser...
-
TopoRetarget: Interaction-Preserving Retargeting for Dexterous Manipulation
TopoRetarget uses a sparse interaction graph and distance-weighted Laplacian deformation optimization with kinematic and penetration constraints to retarget human demonstrations to dexterous hands while preserving tas...
-
LUCID: Learning Embodiment-Agnostic Intent Models from Unstructured Human Videos for Scalable Dexterous Robot Skill Acquisition
LUCID learns embodiment-agnostic intent models from unstructured human videos to train dexterous robot policies in simulation, enabling zero-shot transfer on real-world tasks like stirring and wiping.
-
SynManDex: Synthesizing Human-like Dexterous Grasps from Synthetic Human Pre-Grasps
SynManDex generates human-like dexterous grasps for robots from synthetic human pre-grasps via retargeting and force-closure optimization, reporting 86.4% stability, 4.67/5 human-likeness, 80.7% sim success, and 83.3%...
-
ConTrack: Constrained Hand Motion Tracking with Adaptive Trade-off Control
ConTrack introduces a constrained RL method with online dual-variable adaptation and adaptive resets for improved long-horizon hand tracking in simulation and on real robots.
-
EaDex: A Cross-Embodiment Dexterous Manipulation Framework from Low-Cost Demonstrations
EaDex combines single-camera RGB-D capture, MANO retargeting, and dynamic demonstration annealing to achieve 55.3% relative improvement over baseline on nine cross-embodiment dexterous object-opening tasks across three hands.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.