Learning how to learn: an adaptive dialogue agent for incrementally learning visually grounded word meanings

TL;DR

This paper introduces an adaptive multi-modal dialogue agent trained with reinforcement learning to interactively learn visually grounded word meanings from humans, optimizing for accuracy and minimal human effort.

Contribution

It presents a novel RL-trained dialogue agent capable of incrementally learning visual attributes through natural conversations, outperforming rule-based policies in efficiency.

Findings

The agent effectively learns visual attributes like color and shape.

It balances classifier accuracy and tutoring costs better than rule-based policies.

The system demonstrates coherent interaction with a simulated human tutor.

Abstract

We present an optimised multi-modal dialogue agent for interactive learning of visually grounded word meanings from a human tutor, trained on real human-human tutoring data. Within a life-long interactive learning period, the agent, trained using Reinforcement Learning (RL), must be able to handle natural conversations with human users and achieve good learning performance (accuracy) while minimising human effort in the learning process. We train and evaluate this system in interaction with a simulated human tutor, which is built on the BURCHAK corpus -- a Human-Human Dialogue dataset for the visual learning task. The results show that: 1) The learned policy can coherently interact with the simulated user to achieve the goal of the task (i.e. learning visual attributes of objects, e.g. colour and shape); and 2) it finds a better trade-off between classifier accuracy and tutoring costs than hand-crafted rule-based policies, including ones with dynamic policies.