Generating Grasp Poses for a High-DOF Gripper Using Neural Networks

@article{Liu2019GeneratingGP,
  title={Generating Grasp Poses for a High-DOF Gripper Using Neural Networks},
  author={Min Liu and Zherong Pan and Kai Xu and Kanishka Ganguly and Dinesh Manocha},
  journal={2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  year={2019},
  pages={1518-1525}
}
  • Min Liu, Zherong Pan, +2 authors Dinesh Manocha
  • Published 2019
  • Computer Science, Engineering
  • 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
  • We present a learning-based method for representing grasp poses of a high-DOF hand using neural networks. Due to redundancy in such high-DOF grippers, there exists a large number of equally effective grasp poses for a given target object, making it difficult for the neural network to find consistent grasp poses. We resolve this ambiguity by generating an augmented dataset that covers many possible grasps for each target object and train our neural networks using a consistency loss function to… CONTINUE READING

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 11 CITATIONS

    Deep Differentiable Grasp Planner for High-DOF Grippers

    VIEW 10 EXCERPTS
    CITES METHODS, BACKGROUND & RESULTS

    Multifingered Grasp Planning via Inference in Deep Neural Networks: Outperforming Sampling by Learning Differentiable Models

    VIEW 4 EXCERPTS
    CITES METHODS & BACKGROUND
    HIGHLY INFLUENCED

    6-DOF GraspNet: Variational Grasp Generation for Object Manipulation

    VIEW 1 EXCERPT
    CITES METHODS

    New Formulation of Mixed-Integer Conic Programming for Globally Optimal Grasp Planning

    VIEW 1 EXCERPT
    CITES METHODS

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 39 REFERENCES

    Graspit! A versatile simulator for robotic grasping

    VIEW 8 EXCERPTS
    HIGHLY INFLUENTIAL

    Learning Object Grasping for Soft Robot Hands

    VIEW 9 EXCERPTS
    HIGHLY INFLUENTIAL

    Dex-Net 1.0: A cloud-based network of 3D objects for robust grasp planning using a Multi-Armed Bandit model with correlated rewards

    VIEW 7 EXCERPTS
    HIGHLY INFLUENTIAL

    Dimensionality reduction for hand-independent dexterous robotic grasping

    VIEW 3 EXCERPTS
    HIGHLY INFLUENTIAL

    Learning 6-DOF Grasping Interaction via Deep Geometry-Aware 3D Representations

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    Learning ambidextrous robot grasping policies

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Multi-Task Domain Adaptation for Deep Learning of Instance Grasping from Simulation

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    3D is here: Point Cloud Library (PCL)

    VIEW 1 EXCERPT