DiffusionWorldViewer: Exposing and Broadening the Worldview Reflected by Generative Text-to-Image Models

TL;DR

DiffusionWorldViewer is an interactive tool that reveals and broadens the worldview of generative text-to-image models, helping users understand and align model outputs with diverse human perspectives.

Contribution

The paper introduces DiffusionWorldViewer, a novel interface that exposes TTI model worldviews and enables users to modify outputs to better reflect their perspectives.

Findings

Users can better understand model biases and worldviews.

The tool helps users align generated images with their personal viewpoints.

It broadens the perceived diversity of TTI model outputs.

Abstract

Generative text-to-image (TTI) models produce high-quality images from short textual descriptions and are widely used in academic and creative domains. Like humans, TTI models have a worldview, a conception of the world learned from their training data and task that influences the images they generate for a given prompt. However, the worldviews of TTI models are often hidden from users, making it challenging for users to build intuition about TTI outputs, and they are often misaligned with users' worldviews, resulting in output images that do not match user expectations. In response, we introduce DiffusionWorldViewer, an interactive interface that exposes a TTI model's worldview across output demographics and provides editing tools for aligning output images with user perspectives. In a user study with 18 diverse TTI users, we find that DiffusionWorldViewer helps users represent their varied viewpoints in generated images and challenge the limited worldview reflected in current TTI models.