Timezone: »

Hierarchical Abstraction for Combinatorial Generalization in Object Rearrangement
Michael Chang · Alyssa L Dayan · Franziska Meier · Tom Griffiths · Sergey Levine · Amy Zhang

Object rearrangement is a challenge for embodied agents because solving these tasks requires generalizing across a combinatorially large set of underlying entities that take the value of object states. Worse, these entities are often unknown and must be inferred from sensory percepts. We present a hierarchical abstraction approach to uncover these underlying entities and achieve combinatorial generalization from unstructured inputs. By constructing a factorized transition graph over clusters of object representations inferred from pixels, we show how to learn a correspondence between intervening on states of entities in the agent's model and acting on objects in the environment. We use this correspondence to develop a method for control that generalizes to different numbers and configurations of objects, which outperforms current offline deep RL methods when evaluated on a set of simulated rearrangement and stacking tasks.

Author Information

Michael Chang (University of California, Berkeley)

Ph.D. student at Berkeley AI Research, U.C. Berkeley B.S. in Computer Science from MIT Former research intern under Juergen Schmidhuber, Istituto Dalle Molle di Studi sull'Intelligenza Artificiale (IDSIA) Former undergraduate researcher under Joshua Tenenbaum and Antonio Torralba, MIT

Alyssa L Dayan (University of California, Berkeley)

AI PhD Student at UC Berkeley advised by Stuart Russell

Franziska Meier (Facebook AI Research)
Tom Griffiths (Princeton University)
Sergey Levine (UC Berkeley)
Amy Zhang (Facebook, UC Berkeley)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors