Feature autoencoder for detecting adversarial examples

Ye, Hongwei; Liu, Xiaozhang

doi:10.1002/int.22889

Cited by 3 publications

(4 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dynamic Adversary Training (DAT) was used to strengthen the detector by training the classifier with AEs. In [32], a Feature Autoencoder Detector (FAD) defense framework was developed for detecting AEs. FAD leveraged feature knowledge in the detection process.…”

Section: Related Workmentioning

confidence: 99%

“…Table 6 provides a summary of DL models against AEs for AEDPL-DL and 13 existing methods in the literature , namely BU [19], FS [18], DNR [25], LID [26], ANR [17], FAD [32], EPM [29], SFAD [12], and Regularization [30], against FGSM, PGD, and BIM attacks. The performance metrics include accuracy and AUC.…”

Section: ) Comparison With Existing Modelsmentioning

confidence: 99%

“…Other AE detection methods include steganalysis-based techniques [28], convolutional neural network (CNN)-based approaches with feature regularization, histogram-based feature creation, and residual image reinforcement [29], a data augmentation technique-strengthened binary detector network [30], the feature autoencoder detector framework leveraging feature knowledge [31], and a softmax distribution-based method for detecting misclassified and out-of-distribution examples [32]. However, existing detectors lack robustness, efficiency, and effectiveness when facing new or unknown attacks [11,15].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Framework for Robust Deep Learning Models Against Adversarial Attacks Based on a Protection Layer Approach

Al-Andoli,

Tan,

Sim

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Deep learning (DL) has demonstrated remarkable achievements in various fields. Nevertheless, DL models encounter significant challenges in detecting and defending against adversarial samples (AEs). These AEs are meticulously crafted by adversaries, introducing imperceptible perturbations to clean data to deceive DL models. Consequently, AEs pose potential risks to DL applications. In this paper, we propose an effective framework for enhancing the robustness of DL models against adversarial attacks. The framework leverages convolutional neural networks (CNNs) for feature learning, Deep Neural Networks (DNNs) with softmax for classification, and a defense mechanism to identify and exclude AEs. Evasion attacks are employed to create AEs to evade and mislead the classifier by generating malicious samples during the test phase of DL models i.e., CNN and DNN, using the Fast Gradient Sign Method (FGSM), Basic Iterative Method (BIM), Projected Gradient Descent (PGD), and Square Attack (SA). A protection layer is developed as a detection mechanism placed before the DNN classifier to identify and exclude AEs. The detection mechanism incorporates a machine learning model, which includes one of the following: Fuzzy ARTMAP, Random Forest, K-Nearest Neighbors, XGBoost, or Gradient Boosting Machine. Extensive evaluations are conducted on the MNIST, CIFAR-10, SVHN, and Fashion-MNIST data sets to assess the effectiveness of the proposed framework. The experimental results indicate the framework's ability to effectively and accurately detect AEs generated by four popular attacking methods, highlighting the potential of our developed framework in enhancing its robustness against AEs.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: ) Comparison With Existing Modelsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Framework for Robust Deep Learning Models Against Adversarial Attacks Based on a Protection Layer Approach

Al-Andoli,

Tan,

Sim

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Tong et al [30] integrated Gaussian, mean, and median filtering with AEs for image adversarial detection. Ye et al [31] proposed the FADetector using feature knowledge. In the realm of RF signals, Silvija et al [22] took a statistical approach toward adversarial detection, although with a constrained experimental scope.…”

Section: Related Workmentioning

confidence: 99%

Adversarial Example Detection and Restoration Defensive Framework for Signal Intelligent Recognition Networks

Han,

Qin,

Wang

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

Deep learning-based automatic modulation recognition networks are susceptible to adversarial attacks, posing significant performance vulnerabilities. In response, we introduce a defense framework enriched by tailored autoencoder (AE) techniques. Our design features a detection AE that harnesses reconstruction errors and convolutional neural networks to discern deep features, employing thresholds from reconstruction error and Kullback–Leibler divergence to identify adversarial samples and their origin mechanisms. Additionally, a restoration AE with a multi-layered structure effectively restores adversarial samples generated via optimization methods, ensuring accurate classification. Tested rigorously on the RML2016.10a dataset, our framework proves robust against adversarial threats, presenting a versatile defense solution compatible with various deep learning models.

show abstract

From Attack to Defense

Omar

2024

Innovations, Securities, and Case Studies Across Healthcare, Business, and Technology

View full text Add to dashboard Cite

In recent academic discussions surrounding the textual domain, there has been significant attention directed towards adversarial examples. Despite this focus, the area of detecting such adversarial examples remains notably under-investigated. In this chapter, the authors put forward an innovative approach for the detection of adversarial examples within the realm of natural language processing (NLP). This approach draws inspiration from the local outlier factor (LOF) algorithm. The rigorous empirical evaluation, conducted on pertinent real-world datasets, leverages classifiers based on long short-term memory (LSTM), convolutional neural networks (CNN), and transformer architectures to pinpoint adversarial incursions. The results underscore the superiority of our proposed technique in comparison to recent state-of-the-art methods, namely DISP and FGWS, achieving an impressive F1 detection accuracy rate of up to 94.8%.

show abstract

Feature autoencoder for detecting adversarial examples

Cited by 3 publications

References 39 publications

A Framework for Robust Deep Learning Models Against Adversarial Attacks Based on a Protection Layer Approach

A Framework for Robust Deep Learning Models Against Adversarial Attacks Based on a Protection Layer Approach

Adversarial Example Detection and Restoration Defensive Framework for Signal Intelligent Recognition Networks

From Attack to Defense

Contact Info

Product

Resources

About