Structured Policy Representation: Imposing Stability in arbitrarily conditioned dynamic systems
This work addresses the problem of imposing global stability in arbitrarily conditioned dynamic systems for robotics, which is important for ensuring reasonable behaviors outside demonstration regions.
This paper introduces a new family of deep neural network-based dynamic systems that are globally stable and can be conditioned with an arbitrary context state. These dynamics are demonstrated to be usable as structured robot policies.
We present a new family of deep neural network-based dynamic systems. The presented dynamics are globally stable and can be conditioned with an arbitrary context state. We show how these dynamics can be used as structured robot policies. Global stability is one of the most important and straightforward inductive biases as it allows us to impose reasonable behaviors outside the region of the demonstrations.