mlvlab/SCDM

Official PyTorch implementation of "Stochastic Conditional Diffusion Models for Robust Semantic Image Synthesis" (ICML 2024).

22
/ 100
Experimental

This project helps researchers and engineers generate realistic images from semantic label maps, even when the input labels are imperfect or 'noisy'. You provide an image with labeled regions (like a segmentation mask), and it outputs a high-fidelity, diverse synthetic image that matches those labels. This tool is for anyone working on image synthesis, computer vision, or graphics who needs to create images from abstract semantic information.

No commits in the last 6 months.

Use this if you need to synthesize images from semantic segmentation masks and want robust generation, especially when dealing with potentially imperfect or noisy input labels.

Not ideal if you are looking for a general-purpose image generation tool that doesn't rely on semantic label maps as input, or if your primary goal is not high-fidelity and diverse image outputs.

image-synthesis computer-vision generative-AI semantic-segmentation graphics-research
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

20

Forks

Language

Python

License

MIT

Last pushed

Nov 20, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/mlvlab/SCDM"

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