Abstract: In this work, we aim to learn a unified vision-based policy for multi-fingered robot hands to manipulate a variety of objects in diverse poses. Though prior work has shown benefits of using ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results