yang-song/score_sde

Official code for Score-Based Generative Modeling through Stochastic Differential Equations (ICLR 2021, Oral)

47
/ 100
Emerging

This project offers a unified framework for generating high-quality, realistic images from scratch, or for tasks like image inpainting or colorization. By inputting a noise distribution, it produces diverse and high-fidelity images of specific categories or styles. This tool is ideal for researchers and practitioners in computer vision or machine learning who need to create synthetic datasets, explore generative models, or develop advanced image manipulation techniques.

1,811 stars. No commits in the last 6 months.

Use this if you need to generate new, realistic images (e.g., faces, objects) from noise, or perform conditional image generation tasks like completing missing parts of an image or adding color.

Not ideal if you're looking for an off-the-shelf image editor for quick touch-ups or a system for analyzing existing image data rather than generating new content.

image-generation computer-vision-research synthetic-data image-inpainting deep-learning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 21 / 25

How are scores calculated?

Stars

1,811

Forks

230

Language

Jupyter Notebook

License

Apache-2.0

Last pushed

Nov 29, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/yang-song/score_sde"

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