PICO: Primitive Imitation for COntrol

TL;DR

PICO is a control framework that decomposes demonstrations into primitives, detects novel behaviors, and generalizes to new tasks by blending primitives, demonstrated on robotic platforms.

Contribution

It introduces a novel control method combining imitation learning and task decomposition to identify and generate missing control primitives for new behaviors.

Findings

Successfully detects novel behavior primitives.

Builds missing control policies for new behaviors.

Effective on two robotic platforms.

Abstract

In this work, we explore a novel framework for control of complex systems called Primitive Imitation for Control PICO. The approach combines ideas from imitation learning, task decomposition, and novel task sequencing to generalize from demonstrations to new behaviors. Demonstrations are automatically decomposed into existing or missing sub-behaviors which allows the framework to identify novel behaviors while not duplicating existing behaviors. Generalization to new tasks is achieved through dynamic blending of behavior primitives. We evaluated the approach using demonstrations from two different robotic platforms. The experimental results show that PICO is able to detect the presence of a novel behavior primitive and build the missing control policy.