Generating Exquisite High-Resolution Human-Centric Scenes with Exceptional Text-Image Correspondence Using Pretrained Diffusion Models

BeyondScene, a novel framework, overcomes the limitations of existing text-to-image diffusion models by generating exquisite higher-resolution (over 8K) human-centric scenes with exceptional text-image correspondence and naturalness.