zai-org/ImageReward

[NeurIPS 2023] ImageReward: Learning and Evaluating Human Preferences for Text-to-image Generation

48
/ 100
Emerging

This project helps creators and designers evaluate and refine text-to-image AI generations. It takes a text prompt and a selection of generated images, then provides a score for each image based on human aesthetic preferences. The output helps users select the best images or fine-tune their generation models to produce more appealing visuals.

1,649 stars.

Use this if you need to automatically score and rank images created by text-to-image AI based on what humans generally prefer, or if you want to improve your AI model's ability to generate preferred images.

Not ideal if your primary goal is generating images without any preference-based filtering or if you need an evaluation method based on objective metrics rather than human aesthetic judgments.

AI art generation digital content creation image quality assessment generative AI creative workflows
No Package No Dependents
Maintenance 6 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 16 / 25

How are scores calculated?

Stars

1,649

Forks

89

Language

Python

License

Apache-2.0

Last pushed

Oct 29, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/zai-org/ImageReward"

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