Abstract: Robot manipulation relying on learned object-centric descriptors became popular in recent years. Visual descriptors can easily describe manipulation task objectives, they can be learned ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results