Efficiently Combining Human Demonstrations and Interventions for Safe Training of Autonomous Systems in Real-Time

TL;DR

This paper presents a framework that combines human demonstrations and interventions to train autonomous systems more efficiently and safely in real-time, demonstrated on a quadrotor in simulation.

Contribution

It introduces a Cycle-of-Learning framework that integrates multiple human interaction modalities to enhance training speed and performance of autonomous agents.

Findings

Improved task completion with combined human interaction.

Achieved 32% reduction in data needed for training.

Enhanced safety and efficiency in real-time autonomous training.

Abstract

This paper investigates how to utilize different forms of human interaction to safely train autonomous systems in real-time by learning from both human demonstrations and interventions. We implement two components of the Cycle-of-Learning for Autonomous Systems, which is our framework for combining multiple modalities of human interaction. The current effort employs human demonstrations to teach a desired behavior via imitation learning, then leverages intervention data to correct for undesired behaviors produced by the imitation learner to teach novel tasks to an autonomous agent safely, after only minutes of training. We demonstrate this method in an autonomous perching task using a quadrotor with continuous roll, pitch, yaw, and throttle commands and imagery captured from a downward-facing camera in a high-fidelity simulated environment. Our method improves task completion performance for the same amount of human interaction when compared to learning from demonstrations alone, while also requiring on average 32% less data to achieve that performance. This provides evidence that combining multiple modes of human interaction can increase both the training speed and overall performance of policies for autonomous systems.