Code repository accompanying the paper "On Adversarial Robustness: A Neural Architecture Search perspective"

Overview

Python 3.6

On Adversarial Robustness: A Neural Architecture Search perspective

Preparation:

Clone the repository:

https://github.com/tdchaitanya/nas-robustness.git

prerequisites

  • Python 3.6
  • Pytorch 1.2.0
  • CUDA 10.1

For a hassle-free environment setup, use the environment.yml file included in the repository.

Pre-trained models:

For easy reproduction of the result shown in the paper, this repository is organized dataset-wise, and all the pre-trained models can be downloaded from here

CIFAR-10/100

All the commands in this section should be executed in the cifar directory.

Hand-crafted models on CIFAR-10

All the files corresponding to this dataset are included in cifar-10/100 directories. Download cifar weigths from the shared drive link and place them in nas-robustness/cifar-10/cifar10_models/state_dicts directory.

For running all the four attacks on Resnet-50 (shown in Table 1) run the following command.

python handcrafted.py --arch resnet50

Change the architecture parameter to run attacks on other models. Only resnet-18, resnet-50, densenet-121, densenet-169, vgg-16 are supported for now. For other models, you may have to train them from scratch before running these attacks.

Hand-crafted models on CIFAR-100

For training the models on CIFAR-100 we have used fastai library. Download cifar-100 weigths from the shared drive link and place them in nas-robustness/cifar/c100-weights directory.

Additionally, you'll also have to download the CIFAR-100 dataset from here and place it in the data directory (we'll not be using this anywhere, this is just needed to initialize the fastai model).

python handcrafted_c100.py --arch resnet50
DARTS

Download DARTS CIFAR-10/100 weights from the drive and place it nas-robustness/darts/pretrained

For running all the four attacks on DARTS run the following command:

python darts-nas.py

Add --cifar100 to run the experiments on cifar-100

P-DARTS

Download P-DARTS CIFAR-10/100 weights from the drive and place it nas-robustness/pdarts/pretrained

For running all the four attacks on P-DARTS run the following command:

python pdarts-nas.py

Add --cifar100 to run the experiments on CIFAR-100

NSGA-Net

Download NSGA-Net CIFAR-10/100 weights from the drive and place it nas-robustness/nsga_net/pretrained

For running all the four attacks on P-DARTS run the following command:

python nsganet-nas.py

Add --cifar100 to run the experiments on CIFAR-100

PC-DARTS

Download PC-DARTS CIFAR-10/100 weights from the drive and place it nas-robustness/pcdarts/pretrained

For running all the four attacks on PC-DARTS run the following command:

python pcdarts-nas.py

Add --cifar100 to run the experiments on CIFAR-100

ImageNet

All the commands in this section should be executed in ImageNet directory.

Hand-crafted models

All the files corresponding to this dataset are included in imagenet directory. We use the default pre-trained weights provided by PyTorch for all attacks.

For running all the four attacks on Resnet-50 run the following command:

python handcrafted.py --arch resnet50

For DARTS, P-DARTS, PC-DARTS follow the same instructions as mentioned above for CIFAR-10/100, just change the working directory to ImageNet

DenseNAS

Download DenseNAS ImageNet weights from the drive (these are same as the weights provided in thier official repo) and place it nas-robustness/densenas/pretrained

For running all the four attacks on DenseNAS-R3 run the following command:

python dense-nas.py --model DenseNAS-R3

Citation

@InProceedings{Devaguptapu_2021_ICCV,
    author    = {Devaguptapu, Chaitanya and Agarwal, Devansh and Mittal, Gaurav and Gopalani, Pulkit and Balasubramanian, Vineeth N},
    title     = {On Adversarial Robustness: A Neural Architecture Search Perspective},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops},
    month     = {October},
    year      = {2021},
    pages     = {152-161}
}

Acknowledgements

Some of the code and weights provided in this library are borrowed from the libraries mentioned below:

Owner
Chaitanya Devaguptapu
Masters by Research (M.Tech-RA), IIT Hyderabad
Chaitanya Devaguptapu
This repository contains the code for Direct Molecular Conformation Generation (DMCG).

Direct Molecular Conformation Generation This repository contains the code for Direct Molecular Conformation Generation (DMCG). Dataset Download rdkit

25 Dec 20, 2022
EdiBERT is a generative model based on a bi-directional transformer, suited for image manipulation

EdiBERT, a generative model for image editing EdiBERT is a generative model based on a bi-directional transformer, suited for image manipulation. The

16 Dec 07, 2022
A curated list of the latest breakthroughs in AI (in 2021) by release date with a clear video explanation, link to a more in-depth article, and code.

2021: A Year Full of Amazing AI papers- A Review 📌 A curated list of the latest breakthroughs in AI by release date with a clear video explanation, l

Louis-François Bouchard 2.9k Dec 31, 2022
[NeurIPS 2019] Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss Kaidi Cao, Colin Wei, Adrien Gaidon, Nikos Arechiga, Tengyu Ma This is the offi

Kaidi Cao 528 Jan 01, 2023
State of the Art Neural Networks for Generative Deep Learning

pyradox-generative State of the Art Neural Networks for Generative Deep Learning Table of Contents pyradox-generative Table of Contents Installation U

Ritvik Rastogi 8 Sep 29, 2022
Can we learn gradients by Hamiltonian Neural Networks?

Can we learn gradients by Hamiltonian Neural Networks? This project was carried out as part of the Optimization for Machine Learning course (CS-439) a

2 Aug 22, 2022
A PyTorch-centric hybrid classical-quantum machine learning framework

torchquantum A PyTorch-centric hybrid classical-quantum dynamic neural networks framework. News Add a simple example script using quantum gates to do

MIT HAN Lab 400 Jan 02, 2023
[CVPR 2021] Few-shot 3D Point Cloud Semantic Segmentation

Few-shot 3D Point Cloud Semantic Segmentation Created by Na Zhao from National University of Singapore Introduction This repository contains the PyTor

117 Dec 27, 2022
(CVPR 2021) PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Clouds

PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Clouds by Mutian Xu*, Runyu Ding*, Hengshuang Zhao, and Xiaojuan Qi. Int

CVMI Lab 228 Dec 25, 2022
[PAMI 2020] Show, Match and Segment: Joint Weakly Supervised Learning of Semantic Matching and Object Co-segmentation

Show, Match and Segment: Joint Weakly Supervised Learning of Semantic Matching and Object Co-segmentation This repository contains the source code for

Yun-Chun Chen 60 Nov 25, 2022
A platform for intelligent agent learning based on a 3D open-world FPS game developed by Inspir.AI.

Wilderness Scavenger: 3D Open-World FPS Game AI Challenge This is a platform for intelligent agent learning based on a 3D open-world FPS game develope

46 Nov 24, 2022
Drone detection using YOLOv5

This drone detection system uses YOLOv5 which is a family of object detection architectures and we have trained the model on Drone Dataset. Overview I

Tushar Sarkar 27 Dec 20, 2022
Single Red Blood Cell Hydrodynamic Traps Via the Generative Design

Rbc-traps-generative-design - The generative design for single red clood cell hydrodynamic traps using GEFEST framework

Natural Systems Simulation Lab 4 Jun 16, 2022
A Python library for Deep Graph Networks

PyDGN Wiki Description This is a Python library to easily experiment with Deep Graph Networks (DGNs). It provides automatic management of data splitti

Federico Errica 194 Dec 22, 2022
Nicholas Lee 3 Jan 09, 2022
A whale detector design for the Kaggle whale-detector challenge!

CNN (InceptionV1) + STFT based Whale Detection Algorithm So, this repository is my PyTorch solution for the Kaggle whale-detection challenge. The obje

Tarin Ziyaee 92 Sep 28, 2021
A library for uncertainty quantification based on PyTorch

Torchuq [logo here] TorchUQ is an extensive library for uncertainty quantification (UQ) based on pytorch. TorchUQ currently supports 10 representation

TorchUQ 96 Dec 12, 2022
The world's largest toxicity dataset.

The Toxicity Dataset by Surge AI Saving the internet is fun. Combing through thousands of online comments to build a toxicity dataset isn't. That's wh

Surge AI 134 Dec 19, 2022
Driller: augmenting AFL with symbolic execution!

Driller Driller is an implementation of the driller paper. This implementation was built on top of AFL with angr being used as a symbolic tracer. Dril

Shellphish 791 Jan 06, 2023
Unsupervised Real-World Super-Resolution: A Domain Adaptation Perspective

Unofficial pytorch implementation of the paper "Unsupervised Real-World Super-Resolution: A Domain Adaptation Perspective"

16 Nov 21, 2022