Multi-Modal Dialogue State Tracking for Playing GuessWhich Game

TL;DR

This paper introduces a multi-modal dialogue state tracking model for the GuessWhich game, enabling a Questioner Bot to perform visual reasoning through mental imagery, leading to state-of-the-art results on VisDial datasets.

Contribution

It proposes a novel mental imagery-based dialogue state tracking approach for visual reasoning in GuessWhich, improving over existing methods that lack visual context.

Findings

Achieves new state-of-the-art performance on VisDial datasets.

Effectively models visually related reasoning through mental imagery.

Demonstrates robustness across multiple dataset versions.

Abstract

GuessWhich is an engaging visual dialogue game that involves interaction between a Questioner Bot (QBot) and an Answer Bot (ABot) in the context of image-guessing. In this game, QBot's objective is to locate a concealed image solely through a series of visually related questions posed to ABot. However, effectively modeling visually related reasoning in QBot's decision-making process poses a significant challenge. Current approaches either lack visual information or rely on a single real image sampled at each round as decoding context, both of which are inadequate for visual reasoning. To address this limitation, we propose a novel approach that focuses on visually related reasoning through the use of a mental model of the undisclosed image. Within this framework, QBot learns to represent mental imagery, enabling robust visual reasoning by tracking the dialogue state. The dialogue state comprises a collection of representations of mental imagery, as well as representations of the entities involved in the conversation. At each round, QBot engages in visually related reasoning using the dialogue state to construct an internal representation, generate relevant questions, and update both the dialogue state and internal representation upon receiving an answer. Our experimental results on the VisDial datasets (v0.5, 0.9, and 1.0) demonstrate the effectiveness of our proposed model, as it achieves new state-of-the-art performance across all metrics and datasets, surpassing previous state-of-the-art models. Codes and datasets from our experiments are freely available at \href{https://github.com/xubuvd/GuessWhich}.