Research predicts human preference for text-to-image content before generation

Arxiv CS.CV (computer vision)

Study investigates if human preference for AI-generated images can be predicted prior to generation using diffusion model features. The approach could enable pre-filtering of prompts for higher-quality outputs.

diffusion-modelstext-to-image

Arxiv CS.CV (computer vision)·Joong Ho KimKeith G. Mills

6 days ago