Inspiration Seeds: Learning Non-Literal Visual Combinations for Generative Exploration

TL;DR

Inspiration Seeds introduces a generative framework that creates diverse, coherent visual compositions from two images, fostering open-ended visual exploration without textual prompts, thus aiding early-stage creative ideation.

Contribution

The paper presents a novel, language-free, feed-forward model for generating diverse visual combinations from image pairs, enhancing exploratory visual ideation in creative workflows.

Findings

Produces diverse, coherent image compositions from input pairs

Removes reliance on textual prompts for image generation

Supports fast, intuitive visual recombination for ideation

Abstract

While generative models have become powerful tools for image synthesis, they are typically optimized for executing carefully crafted textual prompts, offering limited support for the open-ended visual exploration that often precedes idea formation. In contrast, designers frequently draw inspiration from loosely connected visual references, seeking emergent connections that spark new ideas. We propose Inspiration Seeds, a generative framework that shifts image generation from final execution to exploratory ideation. Given two input images, our model produces diverse, visually coherent compositions that reveal latent relationships between inputs, without relying on user-specified text prompts. Our approach is feed-forward, trained on synthetic triplets of decomposed visual aspects derived entirely through visual means: we use CLIP Sparse Autoencoders to extract editing directions in CLIP latent space and isolate concept pairs. By removing the reliance on language and enabling fast, intuitive recombination, our method supports visual ideation at the early and ambiguous stages of creative work.