Teaching Text-to-Image Models to Communicate in Dialog

TL;DR

This paper introduces a novel dialog-to-image generation approach that fine-tunes existing text-to-image models to produce high-resolution images aligned with conversational context, improving response quality in dialogue systems.

Contribution

It proposes a tailored fine-tuning method that leverages dialog structure and in-domain data to enhance image generation from conversational context, addressing style mismatch issues.

Findings

Significant improvement over baseline models on PhotoChat and MMDialog Corpus

Effective utilization of dialog structure through linearization and indicators

Enhanced high-resolution image synthesis aligned with dialog context

Abstract

A picture is worth a thousand words, thus, it is crucial for conversational agents to understand, perceive, and effectively respond with pictures. However, we find that directly employing conventional image generation techniques is inadequate for conversational agents to produce image responses effectively. In this paper, we focus on the innovative dialog-to-image generation task, where the model synthesizes a high-resolution image aligned with the given dialog context as a response. To tackle this problem, we design a tailored fine-tuning approach on the top of state-of-the-art text-to-image generation models to fully exploit the structural and semantic features in dialog context during image generation. Concretely, we linearize the dialog context with specific indicators to maintain the dialog structure, and employ in-domain data to alleviate the style mismatch between dialog-to-image and conventional image generation tasks. Empirical results on PhotoChat and MMDialog Corpus show that our approach brings consistent and remarkable improvement with 3 state-of-the-art pre-trained text-to-image generation backbones.