Improving adversarial robustness by a coupling rejection strategy

Last update: Jan 06, 2023

Related tags

Deep Learning Rectified-Rejection

Overview

Adversarial Training with Rectified Rejection

The code for the paper Adversarial Training with Rectified Rejection.

Environment settings and libraries we used in our experiments

This project is tested under the following environment settings:

OS: Ubuntu 18.04.4
GPU: Geforce 2080 Ti or Tesla P100
Cuda: 10.1, Cudnn: v7.6
Python: 3.6
PyTorch: >= 1.6.0
Torchvision: >= 0.6.0

Acknowledgement

The codes are modifed based on Rice et al. 2020, and the model architectures are implemented by pytorch-cifar.

Training Commands

Below we provide running commands training the models with the RR module, taking the setting of PGD-AT + RR (ResNet-18) as an example:

python train_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --attack pgd --lr-schedule piecewise \
                                              --epochs 110 --epsilon 8 \
                                              --attack-iters 10 --pgd-alpha 2 \
                                              --fname auto \
                                              --batch-size 128 \
                                              --adaptivetrain --adaptivetrainlambda 1.0 \
                                              --weight_decay 5e-4 \
                                              --twobranch --useBN \
                                              --selfreweightCalibrate \
                                              --dataset 'CIFAR-10' \
                                              --ATframework 'PGDAT' \
                                              --SGconfidenceW

The FLAG --model_name can be PreActResNet18_twobranch_DenseV1 (ResNet-18) or WideResNet_twobranch_DenseV1 (WRN-34-10). For alternating different AT frameworks, we can set the FLAG --ATframework to be one of PGDAT, TRADES, CCAT.

Evaluation Commands

Below we provide running commands for evaluations.

Evaluating under the PGD attacks

The trained model is saved at trained_models/model_path, where the specific name of model_path is automatically generated during training. The command for evaluating under PGD attacks is:

python eval_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --evalset test --norm l_inf --epsilon 8 \
                                              --attack-iters 1000 --pgd-alpha 2 \
                                              --fname trained_models/model_path \
                                              --load_epoch -1 \
                                              --dataset 'CIFAR-10' \
                                              --twobranch --useBN \
                                              --selfreweightCalibrate

Evaluating under the adaptive CW attacks

The parameter FLAGs --binary_search_steps, --CW_iter, --CW_confidence can be changed, where --detectmetric indicates the rejector that needs to be adaptively evaded.

python eval_cifar_CW.py --model_name PreActResNet18_twobranch_DenseV1 --evalset adaptiveCWtest \
                                              --fname trained_models/model_path \
                                              --load_epoch -1 --seed 2020 \
                                              --binary_search_steps 9 --CW_iter 100 --CW_confidence 0 \
                                              --threatmodel linf --reportmodel linf \
                                              --twobranch --useBN \
                                              --selfreweightCalibrate \
                                              --detectmetric 'RR' \
                                              --dataset 'CIFAR-10'

Evaluating under multi-target and GAMA attacks

The running command for evaluating under multi-target attacks is activated by the FLAG --evalonMultitarget as:

python eval_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --evalset test --norm l_inf --epsilon 8 \
                                              --attack-iters 100 --pgd-alpha 2 \
                                              --fname trained_models/model_path \
                                              --load_epoch -1 \
                                              --dataset 'CIFAR-10' \
                                              --twobranch --useBN \
                                              --selfreweightCalibrate \
                                              --evalonMultitarget --restarts 1

The running command for evaluating under GAMA attacks is activated by the FLAG --evalonGAMA_PGD or --evalonGAMA_FW as:

python eval_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --evalset test --norm l_inf --epsilon 8 \
                                              --attack-iters 100 --pgd-alpha 2 \
                                              --fname trained_models/model_path \
                                              --load_epoch -1 \
                                              --dataset 'CIFAR-10' \
                                              --twobranch --useBN \
                                              --selfreweightCalibrate \
                                              --evalonGAMA_FW

Evaluating under CIFAR-10-C

The running command for evaluating on common corruptions in CIFAR-10-C is:

python eval_cifar_CIFAR10-C.py --model_name PreActResNet18_twobranch_DenseV1 \
                                              --fname trained_models/model_path \
                                              --load_epoch -1 \
                                              --dataset 'CIFAR-10' \
                                              --twobranch --useBN \
                                              --selfreweightCalibrate

Improving adversarial robustness by a coupling rejection strategy

Related tags

Overview

Adversarial Training with Rectified Rejection

Environment settings and libraries we used in our experiments

Acknowledgement

Training Commands

Evaluation Commands

Evaluating under the PGD attacks

Evaluating under the adaptive CW attacks

Evaluating under multi-target and GAMA attacks

Evaluating under CIFAR-10-C

Owner

Tianyu Pang

Official code of ICCV2021 paper "Residual Attention: A Simple but Effective Method for Multi-Label Recognition"

Code & Data for Enhancing Photorealism Enhancement

Optimized primitives for collective multi-GPU communication

Official PyTorch implementation of "Camera Distance-aware Top-down Approach for 3D Multi-person Pose Estimation from a Single RGB Image", ICCV 2019

Wordle Env: A Daily Word Environment for Reinforcement Learning

A series of Python scripts to access measurements from Fluke 28X meters. Fluke IR Remote Interface required.

Official PyTorch Implementation of paper EAN: Event Adaptive Network for Efficient Action Recognition

MediaPipeで姿勢推定を行い、Tokyo2020オリンピック風のピクトグラムを表示するデモ

A Parameter-free Deep Embedded Clustering Method for Single-cell RNA-seq Data

Simple tutorials on Pytorch DDP training

Pytorch implementation of SimSiam Architecture

Official implementation of paper "Query2Label: A Simple Transformer Way to Multi-Label Classification".

Repository of Vision Transformer with Deformable Attention

Pytorch implementation of MixNMatch

Large-scale Hyperspectral Image Clustering Using Contrastive Learning, CIKM 21 Workshop

This is my research project for the Irving Center for Cancer Dynamics/Azizi Lab, Columbia University.

MAGMA - a GPT-style multimodal model that can understand any combination of images and language

The codebase for our paper "Generative Occupancy Fields for 3D Surface-Aware Image Synthesis" (NeurIPS 2021)

STRIVE: Scene Text Replacement In Videos

This is an open source library implementing hyperbox-based machine learning algorithms