UMPNet: Universal Manipulation Policy Network for Articulated Objects

TL;DR

UMPNet is a versatile, image-based policy network capable of manipulating various articulated objects by inferring closed-loop action sequences, supporting 6DoF actions, and generalizing to unseen objects without human demonstrations.

Contribution

The paper introduces UMPNet, a novel universal manipulation policy that learns from diverse objects, uses a new Arrow-of-Time attribute, and operates without human supervision or scripted policies.

Findings

Successfully manipulates a wide range of articulated objects

Generalizes to unseen object categories

Achieves effective multi-step manipulation without human demonstrations

Abstract

We introduce the Universal Manipulation Policy Network (UMPNet) -- a single image-based policy network that infers closed-loop action sequences for manipulating arbitrary articulated objects. To infer a wide range of action trajectories, the policy supports 6DoF action representation and varying trajectory length. To handle a diverse set of objects, the policy learns from objects with different articulation structures and generalizes to unseen objects or categories. The policy is trained with self-guided exploration without any human demonstrations, scripted policy, or pre-defined goal conditions. To support effective multi-step interaction, we introduce a novel Arrow-of-Time action attribute that indicates whether an action will change the object state back to the past or forward into the future. With the Arrow-of-Time inference at each interaction step, the learned policy is able to select actions that consistently lead towards or away from a given state, thereby, enabling both effective state exploration and goal-conditioned manipulation. Video is available at https://youtu.be/KqlvcL9RqKM