We develop an approach for efficient open-vocabulary language-conditioned manipulation policy learning.
We study the problem of action sampling and propose a method to incorporate equivariance properties to the action sampling procedure.
We study sample-efficient modeling of invariant drag coefficients with equivariant networks.
We study how to ground language for robotic grasping while preserve the geometric structure of its symmetry.
We study E(2) Euclidean equivariance in navigation on geometric graphs and develop message passing network to solve it.
We study whether Euclidean symmetry can help in reinforcement learning and planning, which models the geometric transformations between reference frames of robots.
We train an equivariant network for pose prediction from single 2D image by using induced and restricted representations.
We formulate how differentiable planning algorithms can exploit inherent symmetry in path planning problems, named SymPlan, and propose practical algorithms.