An Open-World Simulated Environment for Developmental Robotics

TL;DR

This paper introduces SEDRo, a simulated environment mimicking early human development stages to facilitate self-supervised learning in artificial agents, aiming to improve developmental robotics research.

Contribution

The paper presents a novel simulated environment that models human infant development stages for training and evaluating learning agents in developmental robotics.

Findings

Simulated tests based on developmental psychology are used for evaluation.

SEDRo enables agents to experience developmental stages similar to human infants.

The environment supports self-supervised learning approaches.

Abstract

As the current trend of artificial intelligence is shifting towards self-supervised learning, conventional norms such as highly curated domain-specific data, application-specific learning models, extrinsic reward based learning policies etc. might not provide with the suitable ground for such developments. In this paper, we introduce SEDRo, a Simulated Environment for Developmental Robotics which allows a learning agent to have similar experiences that a human infant goes through from the fetus stage up to 12 months. A series of simulated tests based on developmental psychology will be used to evaluate the progress of a learning model.