This work introduces Joint-Space Empowerment (JoSE), a novel information-theoretic objective that quantifies how much control an agent has over its mechanical degrees-of-freedom, and derives an optimal precoder in closed form under control-affine Gaussian dynamics.
Across multiple dexterous hands, EigenDEXplore consistently outperforms joint-space and learned action-space baselines in grasping, in-hand reorientation, and contact-rich manipulation, and are largest in settings with less reward shaping and curriculum design.
Harsh Gupta, T. Lum, Chang-Hao Wang et al.· 0 citations
Self-motion manifold (SMM) characterizes the geometric structure of the infinite inverse kinematic solutions set of a redundant manipulator at a fixed end-effector pose, and its efficient recovery underpins feasible and global optimal motion planning. Existing methods such as null-space continuation and learning-based...
Continuum robots leverage intrinsic compliance and a high degree of freedom, making them well-suited for interaction in confined spaces. However, their distributed-parameter nature and pronounced nonlinearities pose major challenges for online closed-loop control. Guided by the origin of structural priors and by how co...
Teleoperation is widely used to collect high-quality demonstrations for robot learning. Dexterous manipulation, however, requires coordinated control of motion and contact forces. Operators can specify motion, but struggle to reproduce the subtle force adjustments of human manipulation without feeling the robot's conta...
Ying Yuan, David Held, Jitendra Malik et al.· 0 citations
Robust manipulation under uncertain object dynamics and geometry remains a core challenge in real-world robotics. Policies trained under fixed assumptions often degrade when object mass, friction, scale, or shape vary. Embedding-based adaptation methods such as Rapid Motor Adaptation (RMA) address this by conditioning...
Se June Joo, A. Albini, Luca Wiehe et al.· 0 citations
Hierarchical control architectures are widely used to decompose complex control problems into interacting control levels and are particularly important in robotics, where planning, whole-body motion, and lower-level control must be coordinated across different levels of abstraction and time scales. Their overall closed...
Sebastian Hirt, Lukas Theiner, Jan Peters et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
