MIRACLE: Inverse Reinforcement and Curriculum Learning Model for Human-inspired Mobile Robot Navigation

TL;DR

MIRACLE is a novel inverse reinforcement and curriculum learning framework enabling mobile robots to navigate in emergency scenarios with human-like perception by utilizing gamified stimuli-driven data and deep inverse reinforcement learning.

Contribution

The paper introduces MIRACLE, a new approach combining inverse reinforcement learning and curriculum learning to improve human-like navigation in robots during emergencies.

Findings

Achieved low loss of 2.7717 in a 400-sized environment.

Reduced reliance on demonstrator abilities through deep inverse reinforcement learning.

Enabled robots to interpret stimuli and navigate like humans in emergency scenarios.

Abstract

In emergency scenarios, mobile robots must navigate like humans, interpreting stimuli to locate potential victims rapidly without interfering with first responders. Existing socially-aware navigation algorithms face computational and adaptability challenges. To overcome these, we propose a solution, MIRACLE -- an inverse reinforcement and curriculum learning model, that employs gamified learning to gather stimuli-driven human navigational data. This data is then used to train a Deep Inverse Maximum Entropy Reinforcement Learning model, reducing reliance on demonstrator abilities. Testing reveals a low loss of 2.7717 within a 400-sized environment, signifying human-like response replication. Current databases lack comprehensive stimuli-driven data, necessitating our approach. By doing so, we enable robots to navigate emergency situations with human-like perception, enhancing their life-saving capabilities.