youweiliang/RichHF

Code for CVPR'24 best paper: Rich Human Feedback for Text-to-Image Generation (https://arxiv.org/pdf/2312.10240)

29
/ 100
Experimental

This project helps researchers and developers evaluate the quality of images generated from text descriptions. It takes an image and its corresponding text prompt as input, then generates detailed heatmaps showing areas of 'implausibility' or 'misalignment' and provides objective scores for plausibility, aesthetics, and text-image alignment. It's designed for machine learning researchers working on text-to-image models to understand and improve their creations.

No commits in the last 6 months.

Use this if you are a researcher developing or evaluating text-to-image generation models and need to quantify and visualize how well generated images match human perception and their text prompts.

Not ideal if you need to use this tool for commercial purposes, as both the code and model weights prohibit commercial use.

text-to-image generation image quality assessment generative AI research computer vision machine learning evaluation
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 7 / 25
Maturity 16 / 25
Community 4 / 25

How are scores calculated?

Stars

31

Forks

1

Language

Python

License

Last pushed

Sep 05, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/youweiliang/RichHF"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.