tohinz/semantic-object-accuracy-for-generative-text-to-image-synthesis

Code for "Semantic Object Accuracy for Generative Text-to-Image Synthesis" (TPAMI 2020)

45
/ 100
Emerging

This project helps researchers and developers evaluate how well AI models generate images from text descriptions. It takes a text caption (like "a car is driving down the street") and the image an AI model generates from it. It then checks if the correct objects mentioned in the caption are actually present in the generated image. This tool is designed for AI researchers and practitioners working on generative image models.

106 stars. No commits in the last 6 months.

Use this if you need a reliable and human-correlated metric to assess the quality of images produced by your text-to-image AI model, specifically regarding the presence of described objects.

Not ideal if you are looking for a tool to generate images directly, as this focuses solely on evaluating existing generated images.

AI model evaluation Generative AI Image synthesis Computer vision research Text-to-image
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

106

Forks

23

Language

Python

License

MIT

Category

gan-based-t2i

Last pushed

Jan 13, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/tohinz/semantic-object-accuracy-for-generative-text-to-image-synthesis"

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