Personalizing Text-to-Image Generation via Aesthetic Gradients
This work addresses the problem of customizing AI-generated images for users, but it is incremental as it builds on existing diffusion models.
The authors tackled personalizing text-to-image generation by proposing aesthetic gradients to guide diffusion models toward user-defined aesthetics from image sets, achieving validated results through qualitative and quantitative experiments with stable diffusion and aesthetically-filtered datasets.
This work proposes aesthetic gradients, a method to personalize a CLIP-conditioned diffusion model by guiding the generative process towards custom aesthetics defined by the user from a set of images. The approach is validated with qualitative and quantitative experiments, using the recent stable diffusion model and several aesthetically-filtered datasets. Code is released at https://github.com/vicgalle/stable-diffusion-aesthetic-gradients