DiffusionWorldViewer: Exposing and Broadening the Worldview Reflected by Generative Text-to-Image Models
This addresses the issue for users of text-to-image models who struggle with intuitive understanding and alignment of generated images with their perspectives, though it is incremental as it builds on existing model interfaces.
The paper tackles the problem of hidden and misaligned worldviews in generative text-to-image models, which can lead to outputs that do not match user expectations. It introduces DiffusionWorldViewer, an interactive interface that exposes these worldviews and provides editing tools, and in a user study with 18 users, it helps users represent diverse viewpoints and challenge model limitations.
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.