REVIEW 9 cited by
UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-aware Curriculum and Iterative Generalist-Specialist Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
We propose a novel, object-agnostic method for learning a universal policy for dexterous object grasping from realistic point cloud observations and proprioceptive information under a table-top setting, namely UniDexGrasp++. To address the challenge of learning the vision-based policy across thousands of object instances, we propose Geometry-aware Curriculum Learning (GeoCurriculum) and Geometry-aware iterative Generalist-Specialist Learning (GiGSL) which leverage the geometry feature of the task and significantly improve the generalizability. With our proposed techniques, our final policy shows universal dexterous grasping on thousands of object instances with 85.4% and 78.2% success rate on the train set and test set which outperforms the state-of-the-art baseline UniDexGrasp by 11.7% and 11.3%, respectively.
Forward citations
Cited by 9 Pith papers
-
MANGO-Grasp: Mahalanobis Fields over Geometry-Oriented 3D Gaussians for Cross-Embodiment Dexterous Grasping
MANGO-Grasp uses geometry-oriented 3D Gaussians and Mahalanobis fields to achieve strong cross-embodiment dexterous grasping, with zero-shot transfer to an unseen hand at 84% simulation and 86% real-world success.
-
DextrAH-RGB: Visuomotor Policies to Grasp Anything with Dexterous Hands
A dexterous robot hand learns to grasp novel objects from color images alone, trained purely in simulation, and demonstrates competitive real-world performance versus depth-camera policies.
-
Task-Oriented Human Grasp Synthesis via Context- and Task-Aware Diffusers
A two-stage diffusion framework that learns task-aware contact maps from initial and goal scene point clouds generates human grasps that avoid collisions and complete Placing, Stacking, and Shelving tasks.
-
Crossing the Human-Robot Embodiment Gap with Sim-to-Real RL using One Human Demonstration
A single human RGB-D video, reduced to an object pose trajectory and one pre-manipulation hand pose, suffices to train zero-shot sim-to-real RL policies for dexterous manipulation.
-
COMBO-Grasp: Learning Constraint-Based Manipulation for Bimanual Occluded Grasping
COMBO-Grasp trains a stabilizing constraint policy and an RL grasping policy, then refines the constraint pose with value-function gradients, improving bimanual grasping of occluded objects in simulation and real world.
-
VinT-6D: A Large-Scale Object-in-hand Dataset from Vision, Touch and Proprioception
A new large-scale vision, touch, and proprioception dataset for estimating the 6D pose of objects held in multi-fingered robotic hands, with a baseline network showing improved accuracy when touch is added.
-
BODex: Scalable and Efficient Robotic Dexterous Grasp Synthesis Using Bilevel Optimization
A GPU-parallel bilevel optimization pipeline synthesizes high-quality dexterous grasps faster than prior methods and produces a dataset that improves learned grasping performance from about 40% to 80% in simulation.
-
AdaDexGrasp: Adaptive Dexterous Grasping via 3D Visuo-Tactile Representation Fusion
AdaDexGrasp learns to fuse point clouds with finger-level tactile labels to generate, judge, and correct dexterous grasps, reporting 91%/82%/83% success on seen, unseen-object, and unseen-category sets in simulation.
-
RoboVerse: Towards a Unified Platform, Dataset and Benchmark for Scalable and Generalizable Robot Learning
RoboVerse unifies seven simulators, 15 benchmarks, and 510.5k migrated trajectories into one platform with a four-level generalization benchmark, claiming better robot learning and sim-to-real transfer.
Discussion (0). Continue with ORCID to comment.