adversarial-robustness-toolbox

Adversarial Robustness Toolbox (ART) - Python Library for Machine Learning Security - Evasion, Poisoning, Extraction, Inference - Red and Blue Teams

What is adversarial-robustness-toolbox?

Adversarial Robustness Toolbox (ART) - Python Library for Machine Learning Security - Evasion, Poisoning, Extraction, Inference - Red and Blue Teams

Overview of adversarial-robustness-toolbox

What it is

Adversarial Robustness Toolbox (ART) is a Python library for Machine Learning Security. It is hosted by the Linux Foundation AI & Data Foundation (LF AI & Data). ART provides tools that enable developers and researchers to defend and evaluate Machine Learning models and applications against adversarial threats including Evasion, Poisoning, Extraction, and Inference. The library supports popular machine learning frameworks such as TensorFlow, Keras, PyTorch, scikit-learn, XGBoost, LightGBM, CatBoost, and GPy. It handles all data types (images, tables, audio, video) and machine learning tasks (classification, object detection, speech recognition, generation, certification).

Key features

  • Supports evasion, poisoning, extraction, and inference attacks and defenses.
  • Compatible with multiple ML frameworks: TensorFlow, Keras, PyTorch, scikit-learn, XGBoost, LightGBM, CatBoost, GPy.
  • Works with images, tables, audio, video data types.
  • Covers tasks including classification, object detection, speech recognition, generation, and certification.
  • Provides tools for both red team (attacks) and blue team (defenses) activities.
  • Includes metrics and estimators for evaluating model robustness.

Use cases

ART is used for evaluating and defending machine learning models against adversarial threats. Use cases include testing model resilience to evasion attacks (e.g., adversarial examples), poisoning attacks (data manipulation), model extraction (stealing model functionality), and inference attacks (extracting private information). It is suitable for security researchers, ML developers, and organizations that need to assess and improve the security of their ML systems.

Getting started

See the repository README for installation instructions, examples, and notebooks.

License and pricing

License: MIT. Pricing: open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/trusted-aiadversarial-robustness-toolbox.svg)](https://yunodb.com/tools/trusted-aiadversarial-robustness-toolbox?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/trusted-aiadversarial-robustness-toolbox?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/trusted-aiadversarial-robustness-toolbox.svg" alt="Featured on YunoDB" height="20"></a>