How to Blend Concepts in Diffusion Models

TL;DR

This paper investigates how manipulating latent spaces in diffusion models can enable concept blending, providing insights into effective strategies for combining concepts in generated images.

Contribution

It introduces a method to analyze concept blending in diffusion models by exploring space manipulation and evaluating different text-based combination strategies.

Findings

Concept blending is feasible through latent space manipulation.

The effectiveness of blending strategies depends on the context.

Visual analysis helps evaluate blending quality.

Abstract

For the last decade, there has been a push to use multi-dimensional (latent) spaces to represent concepts; and yet how to manipulate these concepts or reason with them remains largely unclear. Some recent methods exploit multiple latent representations and their connection, making this research question even more entangled. Our goal is to understand how operations in the latent space affect the underlying concepts. To that end, we explore the task of concept blending through diffusion models. Diffusion models are based on a connection between a latent representation of textual prompts and a latent space that enables image reconstruction and generation. This task allows us to try different text-based combination strategies, and evaluate easily through a visual analysis. Our conclusion is that concept blending through space manipulation is possible, although the best strategy depends on the context of the blend.