Lockdown
by git-disl
A backdoor defense for federated learning via isolated subspace training (NeurIPS2023)
AI summary
Federated learning defense system
A backdoor defense system for federated learning, designed to protect against data poisoning attacks by isolating subspace training and aggregating models with robust consensus fusion.
- stars
- 18
- forks
- 3
- watching
- 4
Similar projects
Found by comparing what the projects do, not just their names.
Federated Defense Library
An implementation of a game-theoretic defense against backdoor attacks in federated learning.
Federated Learning Defense
An implementation of a defense against model inversion attacks in federated learning
Federated Learning Attack
This project presents an attack on federated learning systems to compromise their privacy-preserving mechanisms.
Federated attack generator
A tool for demonstrating and analyzing attacks on federated learning systems by introducing backdoors into distributed machine learning models.
Backdoor attacks
This project provides an implementation of backdoor attacks in federated learning frameworks using Python and PyTorch.
Federated Learning Attack
An implementation of a federated learning attack method known as Neurotoxin, which introduces backdoors into machine learning models during the training process.
Attack defense
A defense mechanism against model poisoning attacks in federated learning
Federated Learning Detector
Develops and evaluates a framework for detecting attacks on federated learning systems
Federated Learning System
A PyTorch implementation of an attack-tolerant federated learning system to train robust local models against malicious attacks from adversaries.
Federated Learning Attack
A framework for attacking federated learning systems with adaptive backdoor attacks
Federated Learning Defense
An implementation of a robust federated learning method based on Shapley value to defend against various data and model poisoning attacks
Federated Learning vulnerability research
Researchers investigate vulnerabilities in Federated Learning systems by introducing new backdoor attacks and exploring methods to defend against them.
Gradient protector
A framework designed to protect federated learning models from hijacking attacks by identifying and removing compromised client gradients
Federated Learning Framework
This project presents a framework for robust federated learning against backdoor attacks.
FL defense tool
A toolkit for federated learning with a focus on defending against data heterogeneity
