Recent Advances of Deep Robotic Affordance Learning: A Reinforcement Learning Perspective

TL;DR

This paper reviews recent deep robotic affordance learning (DRAL) developments, focusing on reinforcement learning approaches, their limitations, and future challenges for real-world robotic applications.

Contribution

It classifies DRAL methods from an RL perspective, analyzes technical details, and proposes future directions including predictive affordance modeling.

Findings

RL-based DRAL methods have diverse technical approaches

Current methods face limitations in observation and data collection

Future research should focus on predictive affordance modeling

Abstract

As a popular concept proposed in the field of psychology, affordance has been regarded as one of the important abilities that enable humans to understand and interact with the environment. Briefly, it captures the possibilities and effects of the actions of an agent applied to a specific object or, more generally, a part of the environment. This paper provides a short review of the recent developments of deep robotic affordance learning (DRAL), which aims to develop data-driven methods that use the concept of affordance to aid in robotic tasks. We first classify these papers from a reinforcement learning (RL) perspective, and draw connections between RL and affordances. The technical details of each category are discussed and their limitations identified. We further summarise them and identify future challenges from the aspects of observations, actions, affordance representation, data-collection and real-world deployment. A final remark is given at the end to propose a promising future direction of the RL-based affordance definition to include the predictions of arbitrary action consequences.