Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Utku Ozbulak

Selection of Source Images Heavily Influences the Effectiveness of Adversarial Attacks


Jun 16, 2021
Utku Ozbulak, Esla Timothy Anzaku, Wesley De Neve, Arnout Van Messem


  Access Paper or Ask Questions

Investigating the significance of adversarial attacks and their relation to interpretability for radar-based human activity recognition systems


Jan 26, 2021
Utku Ozbulak, Baptist Vandersmissen, Azarakhsh Jalalvand, Ivo Couckuyt, Arnout Van Messem, Wesley De Neve

* Accepted for publication on Computer Vision and Image Understanding, Special issue on Adversarial Deep Learning in Biometrics & Forensics 

  Access Paper or Ask Questions

Regional Image Perturbation Reduces $L_p$ Norms of Adversarial Examples While Maintaining Model-to-model Transferability


Jul 07, 2020
Utku Ozbulak, Jonathan Peck, Wesley De Neve, Bart Goossens, Yvan Saeys, Arnout Van Messem

* Accepted for the ICML 2020, Workshop on Uncertainty and Robustness in Deep Learning (UDL) 

  Access Paper or Ask Questions

Perturbation Analysis of Gradient-based Adversarial Attacks


Jun 02, 2020
Utku Ozbulak, Manvel Gasparyan, Wesley De Neve, Arnout Van Messem

* Pattern Recognition Letters 2020, Volume 135, Pages 133-120 
* Accepted for publication in Pattern Recognition Letters, 2020 

  Access Paper or Ask Questions

Impact of Adversarial Examples on Deep Learning Models for Biomedical Image Segmentation


Jul 30, 2019
Utku Ozbulak, Arnout Van Messem, Wesley De Neve

* Accepted for the 22nd International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI-19) 

  Access Paper or Ask Questions

Not All Adversarial Examples Require a Complex Defense: Identifying Over-optimized Adversarial Examples with IQR-based Logit Thresholding


Jul 30, 2019
Utku Ozbulak, Arnout Van Messem, Wesley De Neve

* Accepted for the 2019 International Joint Conference on Neural Networks (IJCNN-19) 

  Access Paper or Ask Questions

How the Softmax Output is Misleading for Evaluating the Strength of Adversarial Examples


Nov 21, 2018
Utku Ozbulak, Wesley De Neve, Arnout Van Messem


  Access Paper or Ask Questions