Skill Generalization With Verbs

1Brown University 2California State Polytechnic University 3DeepMind, work done at Brown University

Abstract

It is imperative that robots can understand natural language commands issued by humans. Such commands typically contain verbs that signify what action should be performed on a given object and that are applicable to many objects. We propose a method for generalizing manipulation skills to novel objects using verbs. Our method learns a probabilistic classifier that determines whether a given object trajectory can be described by a specific verb. We show that this classifier accurately generalizes to novel object categories with an average accuracy of 76.69% across 13 object categories and 14 verbs. We then perform policy search over the object kinematics to find an object trajectory that maximizes classifier prediction for a given verb. Our method allows a robot to generate a trajectory for a novel object based on a verb, which can then be used as input to a motion planner. We show that our model can generate trajectories that are usable for executing five verb commands applied to novel instances of two different object categories on a real robot.

Demonstration Video

Dataset

Dataset generated and used for this paper is available upon request.
Please contact the author at rachelm8@mit.edu.

BibTeX

@article{ma2023skill,
      title={{Skill Generalization with Verbs}},
      author={Ma, Rachel and Lam, Lyndon and Spiegel, Benjamin A. and Ganeshan, Aditya and Patel, 
        Roma and Abbatematteo, Ben and Paulius, David and Tellex, Stefanie and Konidaris, George},
      journal={2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
      year={2023},
    }