SunnierLee/DP-ImaGen
[USENIX Security 2024] PrivImage: Differentially Private Synthetic Image Generation using Diffusion Models with Semantic-Aware Pretraining
This tool helps organizations generate realistic synthetic images from their sensitive datasets, like photos of people or confidential documents, without exposing any private information. It takes your private images as input and produces new images that look similar but contain no original identifying data. This is ideal for data scientists, machine learning engineers, and researchers who need to develop and test models using image data while adhering to strict privacy regulations.
No commits in the last 6 months.
Use this if you need to create privacy-preserving synthetic image datasets for research, model training, or sharing, where the original images contain sensitive information.
Not ideal if you need to generate images for creative purposes or if you require an exact replica of your original dataset rather than a privacy-enhanced synthetic version.
Stars
24
Forks
1
Language
Python
License
MIT
Category
Last pushed
Nov 10, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/SunnierLee/DP-ImaGen"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
OPTML-Group/Unlearn-Saliency
[ICLR24 (Spotlight)] "SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in...
Shilin-LU/VINE
[ICLR 2025] "Robust Watermarking Using Generative Priors Against Image Editing: From...
WindVChen/DiffAttack
An unrestricted attack based on diffusion models that can achieve both good transferability and...
koninik/DiffusionPen
Official PyTorch Implementation of "DiffusionPen: Towards Controlling the Style of Handwritten...
Wuyxin/DISC
(ICML 2023) Discover and Cure: Concept-aware Mitigation of Spurious Correlation