---

Documentation | Paper | Samples

[AAAI 2021] DeepRobust is a PyTorch adversarial library for attack and defense methods on images and graphs.

• If you are new to DeepRobust, we highly suggest you read the documentation page or the following content in this README to learn how to use it.
• If you have any questions or suggestions regarding this library, feel free to create an issue here. We will reply as soon as possible :)

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, A Tool and Empirical Studies
• 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

@article{li2020deeprobust,
  title={Deeprobust: A pytorch library for adversarial attacks and defenses},
  author={Li, Yaxin and Jin, Wei and Xu, Han and Tang, Jiliang},
  journal={arXiv preprint arXiv:2005.06149},
  year={2020}
}

Changelog

• [06/2021] [Image Package] Add preprocessing method: APE-GAN.
• [05/2021] DeepRobust is published at AAAI 2021. Check here!
• [05/2021] DeepRobust 0.2.2 Released. Please try pip install deeprobust==0.2.2!
• [04/2021] [Image Package] Add support for ImageNet. See details in test_ImageNet.py
• [04/2021] [Graph Package] Add support for OGB datasets. See more details in the tutorial page.
• [03/2021] [Graph Package] Added node embedding attack and victim models! See this tutorial page.
• [02/2021] [Graph Package] DeepRobust now provides tools for converting the datasets between Pytorch Geometric and DeepRobust. See more details in the tutorial page! DeepRobust now also support GAT, Chebnet and SGC based on pyg; see details in test_gat.py, test_chebnet.py and test_sgc.py
• [12/2020] DeepRobust now can be installed via pip! Try pip install deeprobust!
• [12/2020] [Graph Package] Add four more datasets and one defense algorithm. More details can be found here. More datasets and algorithms will be added later. Stay tuned :)
• [07/2020] Add documentation page!
• [06/2020] Add docstring to both image and graph package

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

Install from pip

pip install deeprobust 

Install from source

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
   from deeprobust.image.utils import download_model
   import torch
   import deeprobust.image.netmodels.resnet as resnet
   from torchvision import transforms,datasets
   
   URL = "https://github.com/I-am-Bot/deeprobust_model/raw/master/CIFAR10_ResNet18_epoch_20.pt"
   download_model(URL, "$MODEL_PATH$")
   
   model = resnet.ResNet18().to('cuda')
   model.load_state_dict(torch.load("$MODEL_PATH$"))
   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, 'cuda')
   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 referred to paper authors' implementations. References can be found at the top of each file.

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

Thanks to their outstanding works!