13 upvotes
Qwen-Image-PE adds design details to vague T2I prompts
What is Qwen-Image-PE: Learning Visual Design for Text-to-Image Prompt Enhancement?
Qwen-Image-PE is a model that expands short, underspecified text prompts into richer visual specifications for text-to-image generators. It learns visual design patterns to keep the original intent while boosting the aesthetic quality of the generated images.
Who it is for
Prompt engineers and AI artists can use it to get more polished images without manually crafting detailed prompts.
How to try it
Open the model card on Hugging Face for a demo