GiorgosKarantonis/Adversarial-Attacks-with-Relativistic-AdvGAN

Using relativism to improve GAN-based Adversarial Attacks. 🦾

41
/ 100
Emerging

This tool helps researchers and security engineers working with image classification models understand and generate 'adversarial examples.' It takes an image dataset and a trained classification model as input, then generates slightly modified versions of those images that can fool the model while still looking normal to a human eye. This is primarily for those exploring the vulnerabilities of deep learning models.

No commits in the last 6 months.

Use this if you need to create visually similar, undetectable changes to images that can trick state-of-the-art image recognition systems, or evaluate the robustness of a trained model against such attacks.

Not ideal if you are looking to defend against adversarial attacks, as this project focuses on generating them rather than building robust models.

AI-security image-recognition deep-learning-vulnerabilities model-robustness adversarial-machine-learning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 17 / 25

How are scores calculated?

Stars

44

Forks

9

Language

Python

License

GPL-3.0

Last pushed

Mar 24, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/GiorgosKarantonis/Adversarial-Attacks-with-Relativistic-AdvGAN"

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