Creative Image Generation with Diffusion Models
This work addresses the problem of generating novel and high-quality images for applications in visual content synthesis, representing an incremental advancement with a new method for a known bottleneck.
The paper tackles creative image generation by proposing a diffusion model framework that drives image probability distributions towards low-probability regions in CLIP embedding space to produce rare and imaginative outputs, achieving high creativity without sacrificing visual fidelity.
Creative image generation has emerged as a compelling area of research, driven by the need to produce novel and high-quality images that expand the boundaries of imagination. In this work, we propose a novel framework for creative generation using diffusion models, where creativity is associated with the inverse probability of an image's existence in the CLIP embedding space. Unlike prior approaches that rely on a manual blending of concepts or exclusion of subcategories, our method calculates the probability distribution of generated images and drives it towards low-probability regions to produce rare, imaginative, and visually captivating outputs. We also introduce pullback mechanisms, achieving high creativity without sacrificing visual fidelity. Extensive experiments on text-to-image diffusion models demonstrate the effectiveness and efficiency of our creative generation framework, showcasing its ability to produce unique, novel, and thought-provoking images. This work provides a new perspective on creativity in generative models, offering a principled method to foster innovation in visual content synthesis.