caiyuanhao1998/PNGAN

"Learning to Generate Realistic Noisy Images via Pixel-level Noise-aware Adversarial Training" (NeurIPS 2021)

44
/ 100
Emerging

This project helps image processing researchers and developers create highly realistic noisy versions of clean images. By taking a clean image, it generates a synthetic noisy image that closely mimics real-world sensor noise. The end user is typically a researcher or engineer working on developing and testing image denoising algorithms.

146 stars. No commits in the last 6 months.

Use this if you need to generate high-fidelity, synthetic noisy images to train and evaluate image denoising models, especially when real noisy-clean image datasets are scarce or expensive to acquire.

Not ideal if you primarily need to remove noise from existing noisy images, as this tool focuses on generating realistic noise rather than directly performing denoising.

image-denoising computer-vision-research image-quality-testing synthetic-data-generation digital-photography
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 16 / 25

How are scores calculated?

Stars

146

Forks

19

Language

Python

License

MIT

Last pushed

Jun 13, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/caiyuanhao1998/PNGAN"

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