jonasricker/aeroblade

[CVPR2024] AEROBLADE: Training-Free Detection of Latent Diffusion Images Using Autoencoder Reconstruction Error

31
/ 100
Emerging

This tool helps you determine if a digital image was created by a latent diffusion model like Stable Diffusion or Kandinsky, rather than being a real photograph. You provide an image or a folder of images, and it outputs a score indicating the likelihood of each image being AI-generated. This is useful for content moderators, journalists, or anyone needing to verify the authenticity of digital images.

No commits in the last 6 months.

Use this if you need to quickly assess whether an image or a collection of images was generated by popular AI diffusion models, without requiring any prior training.

Not ideal if you need to detect AI-generated images from models other than Stable Diffusion 1/2 or Kandinsky 2.1, or if you need to detect deepfakes or other forms of image manipulation.

content-moderation image-verification digital-forensics journalism AI-content-detection
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 8 / 25
Community 15 / 25

How are scores calculated?

Stars

69

Forks

11

Language

Python

License

Last pushed

Dec 09, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/jonasricker/aeroblade"

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