Please wait ...
0% Complete
Home
/
13th International Conference on Computer and Knowledge Engineering
A large input-space-margin approach for adversarial training
Authors :
Reihaneh Nikouei
1
Mohammad Taheri
2
1- Department of Computer Science and Engineering, Shiraz University, Shiraz, Iran.
2- Department of Computer Science and Engineering, Shiraz University, Shiraz, Iran.
Keywords :
Adversarial attack،Large margin،Input space،Defense method
Abstract :
It is shown that machine learning models are vulnerable to adversarial attacks. Therefore, different defense methods such as adversarial training have been proposed to improve models’ robustness against these attacks. Some recent approaches proposed to structurally improve the robustness of the models. For example, large margin methods try to increase a margin, empty of instances, along decision boundaries that structurally increase necessary change to modify a training instance to an adversarial one. However, nonlinear large-margin models, maximize the margin in a high dimensional space although adversarial examples are generated with a little change in the original space. In this paper, a novel mixed approach is proposed, called LIM (Large Input Margin) to improve the robustness of the model by minimizing both structural and empirical risks. Specifically, both training and adversarial example generation are done based on a loss function to maximize the margin in the original feature space even in a non-linear model. The proposed method is evaluated with FGSM and PGD attacks on MNIST and CIFAR10 datasets. The experimental results show that LIM method outperforms the state-of-the-art defense methods significantly and improves adversarial robustness against FGSM and PGD attacks on both datasets.
Papers List
List of archived papers
An Evolutionary Approach with Surrogate Models for Feature Selection in Intrusion Detection Systems
Sadeq Moradi - Hadi Shahriar Shahhoseini
Distilling Knowledge from CNN-Transformer Models for Enhanced Human Action Recognition
Hamid Ahmadabadi - Omid Nejati Manzari - Ahmad Ayatollahi
An effective hybrid algorithm for locating splicing forgery image
Seyed Hesamoddin Hosseini - Amene Vatanparast - Amir Hossein Taherinia
Improving ADHD Detection with Cost-Sensitive LightGBM
Behnam Yousefimehr - Mehdi Ghatee - Ali Heydari
Realism in Action: Anomaly-Aware Diagnosis of Brain Tumors from Medical Images Using YOLOv8 and DeiT
Seyed Mohammad Hossein Hashemi - Leila Safari - Mohsen Hooshmand - Amirhossein Dadashzadeh Taromi
DIPT: Diversified Personalized Transformer for QAC systems
Mahdi Dehghani - Samira Vaez Barenji - Saeed Farzi
Robust Learning to Learn Graph Topologies
Navid Akhavan Attar - Ali Fahim
Swin-RSCBNet: A Transformer-Based Network for Skin Cancer Segmentation with Multi-Scale and Attention Modules
Benyamin Mirab Golkhatmi - Mostafa Heydari - Mahboobeh Houshmand - Seyyed Abed Hosseini
Distinguishing Abstracts of Human-Written and ChatGPT-Generated Papers in the Field of Computer Science
Mohsen Arzani - Hamed Vahdat-Nejad - Matin Hossein-Pour
Effect of Tissue Excitation in Breast Cancer Detection from Ultrasound RF Time Series: Phantom studies
Elaheh Norouzi Ghehi - Ali Fallah - Saeid Rashidi - Maryam Mehdizadeh Dastjerdi
more
Samin Hamayesh - Version 46.0.2