Universal Adversarial Perturbations (UAPs) for PyTorch

Official implementation of the ICCV2023 paper: Enhancing Generalization of Universal Adversarial Perturbation through Gradient Aggregation

Feature Norm Supressing against Universal Patch Attack

Satellite images classification

Generative Adversarial Perturbations (CVPR 2018)

TRADES (TRadeoff-inspired Adversarial DEfense via Surrogate-loss minimization)

Reproduce work "LaVAN: Localized and Visible Adversarial Noise - ICML2018"

Code/Models for Defending Against Universal Attacks Through Selective Feature Regeneration, CVPR 2020

Convert Caffe models to PyTorch

Caffe: a fast open framework for deep learning.

This is the implementation for IEEE S&P 2022 paper "Model Orthogonalization: Class Distance Hardening in Neural Networks for Better Security."

Codes for NeurIPS 2021 paper "Adversarial Neuron Pruning Purifies Backdoored Deep Models"

This is an implementation demo of the ICLR 2021 paper [Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks](https://openreview.net/pdf?id=9l0K4OM-oXE) in PyTorch.

Provable adversarial robustness at ImageNet scale

Influence Estimation for Gradient-Boosted Decision Trees

Source code of paper "FG-UAP: Feature-Gathering Universal Adversarial Perturbation"

The official implementation of the CCS'23 paper, Narcissus clean-label backdoor attack -- only takes THREE images to poison a face recognition dataset in a clean-label way and achieves a 99.89% att…

Official repository for "On Generating Transferable Targeted Perturbations" (ICCV 2021)

A set of examples around pytorch in Vision, Text, Reinforcement Learning, etc.

Code for "Label-Consistent Backdoor Attacks"