Many Episode Learning in a Modular Embodied Agent via End-to-End Interaction

TL;DR

This paper presents a modular embodied agent that improves through end-to-end interactions with crowd-workers, utilizing a combination of learned and heuristic modules, and a credit assignment system for continuous learning.

Contribution

It introduces a novel framework for self-improving embodied agents via crowd-sourced annotations and credit assignment in an end-to-end interaction setting.

Findings

Agent performance improved over multiple interaction rounds.

Effective credit assignment enabled targeted module updates.

Crowd-sourcing facilitated continuous learning and adaptation.

Abstract

In this work we give a case study of an embodied machine-learning (ML) powered agent that improves itself via interactions with crowd-workers. The agent consists of a set of modules, some of which are learned, and others heuristic. While the agent is not "end-to-end" in the ML sense, end-to-end interaction is a vital part of the agent's learning mechanism. We describe how the design of the agent works together with the design of multiple annotation interfaces to allow crowd-workers to assign credit to module errors from end-to-end interactions, and to label data for individual modules. Over multiple automated human-agent interaction, credit assignment, data annotation, and model re-training and re-deployment, rounds we demonstrate agent improvement.