Investigating Conceptual Blending of a Diffusion Model for Improving Nonword-to-Image Generation

TL;DR

This paper analyzes how diffusion models blend concepts in nonword-to-image generation, proposing a new embedding conversion method that enhances image quality and conceptual blending for more intuitive and creative outputs.

Contribution

It introduces a quantitative analysis of conceptual blending in diffusion models and proposes an improved embedding space conversion method for better nonword-to-image generation.

Findings

High percentage of blended concepts in generated images.

Proposed embedding conversion improves image quality.

Blending mainly occurs in specific high-dimensional embedding dimensions.

Abstract

Text-to-image diffusion models sometimes depict blended concepts in the generated images. One promising use case of this effect would be the nonword-to-image generation task which attempts to generate images intuitively imaginable from a non-existing word (nonword). To realize nonword-to-image generation, an existing study focused on associating nonwords with similar-sounding words. Since each nonword can have multiple similar-sounding words, generating images containing their blended concepts would increase intuitiveness, facilitating creative activities and promoting computational psycholinguistics. Nevertheless, no existing study has quantitatively evaluated this effect in either diffusion models or the nonword-to-image generation paradigm. Therefore, this paper first analyzes the conceptual blending in a pretrained diffusion model, Stable Diffusion. The analysis reveals that a high percentage of generated images depict blended concepts when inputting an embedding interpolating between the text embeddings of two text prompts referring to different concepts. Next, this paper explores the best text embedding space conversion method of an existing nonword-to-image generation framework to ensure both the occurrence of conceptual blending and image generation quality. We compare the conventional direct prediction approach with the proposed method that combines $k$-nearest neighbor search and linear regression. Evaluation reveals that the enhanced accuracy of the embedding space conversion by the proposed method improves the image generation quality, while the emergence of conceptual blending could be attributed mainly to the specific dimensions of the high-dimensional text embedding space.