Enhancing Creative Generation on Stable Diffusion-based Models
This addresses a specific limitation in generative AI for users seeking more creative image outputs, though it appears incremental as it builds on existing models without major architectural changes.
The paper tackles the problem of limited creative capability in Stable Diffusion-based text-to-image models by introducing C3 (Creative Concept Catalyst), a training-free approach that selectively amplifies features during denoising to foster more creative outputs, demonstrating effectiveness across various models.
Recent text-to-image generative models, particularly Stable Diffusion and its distilled variants, have achieved impressive fidelity and strong text-image alignment. However, their creative capability remains constrained, as including `creative' in prompts seldom yields the desired results. This paper introduces C3 (Creative Concept Catalyst), a training-free approach designed to enhance creativity in Stable Diffusion-based models. C3 selectively amplifies features during the denoising process to foster more creative outputs. We offer practical guidelines for choosing amplification factors based on two main aspects of creativity. C3 is the first study to enhance creativity in diffusion models without extensive computational costs. We demonstrate its effectiveness across various Stable Diffusion-based models.