pmh9960/GCDP

Official PyTorch implementation of "Learning to Generate Semantic Layouts for Higher Text-Image Correspondence in Text-to-Image Synthesis." (ICCV 2023)

21
/ 100
Experimental

This project helps graphic designers, content creators, or researchers who need to generate realistic images from text descriptions, especially when working with specialized datasets like urban scenes or celebrity faces. It takes a text prompt and generates both a high-quality image and a corresponding 'semantic layout' (like a mask showing where objects are). This ensures the generated image accurately matches the text description, even with limited training data.

No commits in the last 6 months.

Use this if you need to generate highly accurate images from text descriptions within specific domains where large text-image datasets are not available, such as for architectural visualizations or character design.

Not ideal if you are working with generic image generation where web-scale datasets are abundant, as its primary benefit is in improving correspondence for niche domains.

Image Generation Content Creation Computer Vision Research Digital Art Semantic Segmentation
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 8 / 25
Community 5 / 25

How are scores calculated?

Stars

46

Forks

2

Language

Python

License

Last pushed

Nov 02, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/pmh9960/GCDP"

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