Task-Oriented Dexterous Grasp Synthesis
Using Differentiable Grasp Wrench Boundary Estimator

IROS 2024


Jiayi Chen1,2    Yuxing Chen1,3    Jialiang Zhang1,3    He Wang1,2†

1Peking University    2Beijing Academy of Artificial Intelligence    3Galbot   

corresponding author  


input

Synthesized task-oriented dexterous grasps. GWS is expected to cover TWS for each task.

Grasp Wrench Space (GWS) : all wrenches that can be applied on object through hand contacts.
Task Wrench Space (TWS) : all wrenches that should be applied on object during task execution. TWS is approximated as a 6D hyper-fan and given as a task prior in this work.


Abstract


This work tackles the problem of task-oriented dexterous hand pose synthesis, which involves generating a static hand pose capable of applying a task-specific set of wrenches to manipulate objects. Unlike previous approaches that focus solely on force-closure grasps, which are unsuitable for non-prehensile manipulation tasks (e.g., turning a knob or pressing a button), we introduce a unified framework covering force-closure grasps, non-force-closure grasps, and a variety of non-prehensile poses. Our key idea is a novel optimization objective quantifying the disparity between the Task Wrench Space (TWS, the desired wrenches predefined as a task prior) and the Grasp Wrench Space (GWS, the achievable wrenches computed from the current hand pose). By minimizing this objective, gradient-based optimization algorithms can synthesize task-oriented hand poses without additional human demonstrations. Our specific contributions include 1) a fast, accurate, and differentiable technique for estimating the GWS boundary; 2) a task-oriented objective function based on the disparity between the estimated GWS boundary and the provided TWS boundary; and 3) an efficient implementation of the synthesis pipeline that leverages CUDA accelerations and supports largescale paralleling. Experimental results on 10 diverse tasks demonstrate a 72.6% success rate in simulation. Furthermore, real-world validation for 4 tasks confirms the effectiveness of synthesized poses for manipulation. Notably, despite being primarily tailored for task-oriented hand pose synthesis, our pipeline can generate force-closure grasps 50 times faster than DexGraspNet while maintaining comparable grasp quality.


Video




Real-World Experiments




Gradient-based Optimization Process


Rotate Knob
(specify torque)

Turn Handle
(specify force)

Turn Handle
(force closure)

Grasp Toy
(force closure)



Qualitative Results (different tasks)


Rotate Key

Rotate Lid

Rotate Knob

Pull Closet

Drag Chair

Pinch Cable

Grasp Mug

Lift Handbag

Lift Plate

Lift Suitcase

Press Stapler

Press Button



Qualitative Results (same task)


Rotate Three-Lobed Knob



Citation


@article{chen2023task,
  title={Task-Oriented Dexterous Grasp Synthesis via Differentiable Grasp Wrench Boundary Estimator},
  author={Chen, Jiayi and Chen, Yuxing and Zhang, Jialiang and Wang, He},
  journal={arXiv preprint arXiv:2309.13586},
  year={2023}
}

Contact


If you have any questions, please feel free to contact Jiayi Chen at jiayichen@pku.edu.cn, and He Wang at hewang@pku.edu.cn.