Skip to content
#

image-text-alignment-evaluation

Here is 1 public repository matching this topic...

ImageFusion blends people, background, and props from multiple images into a single, realistic scene using OpenAI’s image generation API. It uses advanced prompt summarization and engineering, with output evaluated via CLIP-based similarity for alignment and trustworthiness.

  • Updated Apr 29, 2025
  • Python

Add this topic to your repo

To associate your repository with the image-text-alignment-evaluation topic, visit your repo's landing page and select "manage topics."

Learn more