Accuracy of TextFooler black box adversarial attacks on 01 loss sign activation neural network ensemble (2024-02-12T00:00:00.000000Z)

TL;DR

The work here suggests that 01 loss sign activation networks could be further developed to create fool proof models against text adversarial attacks.

Abstract

Recent work has shown the defense of 01 loss sign activation neural networks against image classification adversarial attacks. A public challenge to attack the models on CIFAR10 dataset remains undefeated. We ask the following question in this study: are 01 loss sign activation neural networks hard to deceive with a popular black box text adversarial attack program called TextFooler? We study this question on four popular text classification datasets: IMDB reviews, Yelp reviews, MR sentiment classification, and AG news classification. We find that our 01 loss sign activation network is much harder to attack with TextFooler compared to sigmoid activation cross entropy and binary neural networks. We also study a 01 loss sign activation convolutional neural network with a novel global pooling step specific to sign activation networks. With this new variation we see a significant gain in adversarial accuracy rendering TextFooler practically useless against it. We make our code freely available at \url{https://github.com/zero-one-loss/wordcnn01} and \url{https://github.com/xyzacademic/mlp01example}. Our work here suggests that 01 loss sign activation networks could be further developed to create fool proof models against text adversarial attacks.

Authors

Yunzhe Xue

1 papers

Usman Roshan

1 papers

Accuracy of TextFooler black box adversarial attacks on 01 loss sign activation neural network ensemble

TL;DR

Abstract

Authors

References22 items

Why adversarial training can hurt robust accuracy

Defense Against Adversarial Attacks Based on Stochastic Descent Sign Activation Networks on Medical Images

Accurate and adversarially robust classification of medical images and ECG time-series with gradient-free trained sign activation neural networks

Towards adversarial robustness with 01 loss neural networks

On the transferability of adversarial examples between convex and 01 loss models

Larq: An Open-Source Library for Training Binarized Neural Networks

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Stochastic Coordinate Descent for 01 Loss and Its Sensitivity to Adversarial Attacks

Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment

Adversarial Training Can Hurt Generalization

Bi-Real Net: Enhancing the Performance of 1-bit CNNs With Improved Representational Capability and Advanced Training Algorithm

Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples

Adversarial Machine Learning at Scale

Support vector machines and Word2vec for text classification with semantic features

GloVe: Global Vectors for Word Representation

Convolutional Neural Networks for Sentence Classification

Intriguing properties of neural networks

Scikit-learn: Machine Learning in Python

Accuracy of white box and black box adversarial attacks on a sign activation 01 loss neural network ensemble

Binarized Neural Networks

Random Forests

Operational Feasibility of Adversarial Attacks Against Artificial Intelligence

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names