









Instruct Pix2Pix Diffusion is a method for editing images using natural language instructions. It leverages a pre-trained text-to-image diffusion model and trains a lightweight instruction-following model on synthetic data. This allows users to edit images by simply providing textual instructions, such as "make the sky more blue" or "add a cat to the image."
Users provide an image and a text instruction. The Instruct Pix2Pix model then modifies the image according to the instruction, generating an edited version.



37.10%
