Autonomous Learning of Object-Centric Abstractions for High-Level Planning

We propose a method for autonomously learning an object-centric representation of a highdimensional environment that is suitable for planning. Such abstractions can be immediately transferred between tasks that share the same types of objects, resulting in agents that require fewer samples to learn a model of a new task. We demonstrate our approach on a series of Minecraft tasks to learn object-centric representations-directly from pixel data-that can be leveraged to quickly solve new tasks. The resulting learned representations enable the use of a task-level planner, resulting in an agent capable of forming complex, long-term plans. 1 Recent work has shown how to learn an abstraction of a task that is provably suitable for planning with a given set of high-level actions (Konidaris et al., 2018) . However, these representations are highly task-specific and must be relearned for any new task, or even any small change to an existing task. This makes them fatally impractical, especially for an agent that must solve multiple complex tasks.