DeepRobust

DeepRobust is a pytorch adversarial library for attack and defense methods on images and graphs.

List of including algorithms can be found in [Image Package] and [Graph Package].

Environment & Installation

Usage

• Image Attack and Defense

• Graph Attack and Defense

Acknowledgement

For more details about attacks and defenses, you can read the following papers.

Adversarial Attacks and Defenses on Graphs: A Review and Empirical Study

Adversarial Attacks and Defenses in Images, Graphs and Text: A Review

If our work could help your research, please cite: DeepRobust: A PyTorch Library for Adversarial Attacks and Defenses

Basic Environment

• python >= 3.6 (python 3.5 should also work)
• pytorch >= 1.2.0

see setup.py or requirements.txt for more information.

Installation

git clone https://github.com/DSE-MSU/DeepRobust.git
cd DeepRobust
python setup.py install

Test Examples

python examples/image/test_PGD.py
python examples/image/test_pgdtraining.py
python examples/graph/test_gcn_jaccard.py --dataset cora
python examples/graph/test_mettack.py --dataset cora --ptb_rate 0.05

Usage

Image Attack and Defense

1. Train model

   Example: Train a simple CNN model on MNIST dataset for 20 epoch on gpu.

   import deeprobust.image.netmodels.train_model as trainmodel
   trainmodel.train('CNN', 'MNIST', 'cuda', 20)

   Model would be saved in deeprobust/trained_models/.

2. Instantiated attack methods and defense methods.

   Example: Generate adversary example with PGD attack.

   from deeprobust.image.attack.pgd import PGD
   from deeprobust.image.config import attack_params
   import torch
   import deeprobust.image.netmodels.resnet as resnet
   
   model = resnet.ResNet18().to('cuda')
   model.load_state_dict(torch.load("./trained_models/CIFAR10_ResNet18_epoch_50.pt"))
   model.eval()
   
   transform_val = transforms.Compose([transforms.ToTensor()])
   test_loader  = torch.utils.data.DataLoader(
                   datasets.CIFAR10('deeprobust/image/data', train = False, download=True,
                   transform = transform_val),
                   batch_size = 10, shuffle=True)
   
   x, y = next(iter(test_loader))
   x = x.to('cuda').float()
   
   adversary = PGD(model, device)
   Adv_img = adversary.generate(x, y, **attack_params['PGD_CIFAR10'])

   Example: Train defense model.

   from deeprobust.image.defense.pgdtraining import PGDtraining
   from deeprobust.image.config import defense_params
   from deeprobust.image.netmodels.CNN import Net
   import torch
   from torchvision import datasets, transforms 
   
   model = Net()
   train_loader = torch.utils.data.DataLoader(
                   datasets.MNIST('deeprobust/image/defense/data', train=True, download=True,
                                   transform=transforms.Compose([transforms.ToTensor()])),
                                   batch_size=100,shuffle=True)
   
   test_loader = torch.utils.data.DataLoader(
                 datasets.MNIST('deeprobust/image/defense/data', train=False,
                               transform=transforms.Compose([transforms.ToTensor()])),
                               batch_size=1000,shuffle=True)
   
   defense = PGDtraining(model, 'cuda')
   defense.generate(train_loader, test_loader, **defense_params["PGDtraining_MNIST"])

   More example code can be found in deeprobust/examples.

3. Use our evulation program to test attack algorithm against defense.

   Example:

   cd DeepRobust
   python examples/image/test_train.py
   python deeprobust/image/evaluation_attack.py
   

Graph Attack and Defense

Attacking Graph Neural Networks

1. Load dataset

   import torch
   import numpy as np
   from deeprobust.graph.data import Dataset
   from deeprobust.graph.defense import GCN
   from deeprobust.graph.global_attack import Metattack
   
   data = Dataset(root='/tmp/', name='cora', setting='nettack')
   adj, features, labels = data.adj, data.features, data.labels
   idx_train, idx_val, idx_test = data.idx_train, data.idx_val, data.idx_test
   idx_unlabeled = np.union1d(idx_val, idx_test)

2. Set up surrogate model

   device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
   surrogate = GCN(nfeat=features.shape[1], nclass=labels.max().item()+1, nhid=16,
                   with_relu=False, device=device)
   surrogate = surrogate.to(device)
   surrogate.fit(features, adj, labels, idx_train)

3. Set up attack model and generate perturbations

   model = Metattack(model=surrogate, nnodes=adj.shape[0], feature_shape=features.shape, device=device)
   model = model.to(device)
   perturbations = int(0.05 * (adj.sum() // 2))
   model.attack(features, adj, labels, idx_train, idx_unlabeled, perturbations, ll_constraint=False)
   modified_adj = model.modified_adj

For more details please refer to mettack.py or run python examples/graph/test_mettack.py --dataset cora --ptb_rate 0.05

Defending Against Graph Attacks

1. Load dataset

   import torch
   from deeprobust.graph.data import Dataset, PtbDataset
   from deeprobust.graph.defense import GCN, GCNJaccard
   import numpy as np
   np.random.seed(15)
   
   # load clean graph
   data = Dataset(root='/tmp/', name='cora', setting='nettack')
   adj, features, labels = data.adj, data.features, data.labels
   idx_train, idx_val, idx_test = data.idx_train, data.idx_val, data.idx_test
   
   # load pre-attacked graph by mettack
   perturbed_data = PtbDataset(root='/tmp/', name='cora')
   perturbed_adj = perturbed_data.adj

2. Test

   # Set up defense model and test performance
   device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
   model = GCNJaccard(nfeat=features.shape[1], nclass=labels.max()+1, nhid=16, device=device)
   model = model.to(device)
   model.fit(features, perturbed_adj, labels, idx_train)
   model.eval()
   output = model.test(idx_test)
   
   # Test on GCN
   model = GCN(nfeat=features.shape[1], nclass=labels.max()+1, nhid=16, device=device)
   model = model.to(device)
   model.fit(features, perturbed_adj, labels, idx_train)
   model.eval()
   output = model.test(idx_test)

For more details please refer to test_gcn_jaccard.py or run python examples/graph/test_gcn_jaccard.py --dataset cora

Sample Results

adversary examples generated by fgsm:

Left:original, classified as 6; Right:adversary, classified as 4.

Serveral trained models can be found here: https://drive.google.com/open?id=1uGLiuCyd8zCAQ8tPz9DDUQH6zm-C4tEL

Acknowledgement

Some of the algorithms are refer to paper authors' implementations. References can be found at the top of each file.

Implementation of network structure are refer to weiaicunzai's github. Original code can be found here: pytorch-cifar100

Thanks to their outstanding works!