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 Seyed-Mohsen Moosavi-Dezfooli

Seyed-Mohsen Moosavi-Dezfooli

Are socially-aware trajectory prediction models really socially-aware?


Aug 24, 2021
Saeed Saadatnejad, Mohammadhossein Bahari, Pedram Khorsandi, Mohammad Saneian, Seyed-Mohsen Moosavi-Dezfooli, Alexandre Alahi


  Access Paper or Ask Questions

Adversarial training may be a double-edged sword


Jul 24, 2021
Ali Rahmati, Seyed-Mohsen Moosavi-Dezfooli, Huaiyu Dai

* Presented as a RobustML workshop paper at ICLR 2021 

  Access Paper or Ask Questions

What can linearized neural networks actually say about generalization?


Jun 12, 2021
Guillermo Ortiz-Jiménez, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard

* 17 pages, 15 figures 

  Access Paper or Ask Questions

Uniform Convergence, Adversarial Spheres and a Simple Remedy


May 07, 2021
Gregor Bachmann, Seyed-Mohsen Moosavi-Dezfooli, Thomas Hofmann


  Access Paper or Ask Questions

Understanding Catastrophic Overfitting in Adversarial Training


May 06, 2021
Peilin Kang, Seyed-Mohsen Moosavi-Dezfooli


  Access Paper or Ask Questions

A neural anisotropic view of underspecification in deep learning


Apr 29, 2021
Guillermo Ortiz-Jimenez, Itamar Franco Salazar-Reque, Apostolos Modas, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard

* Presented as a RobustML workshop paper at ICLR 2021 

  Access Paper or Ask Questions

Optimism in the Face of Adversity: Understanding and Improving Deep Learning through Adversarial Robustness


Oct 19, 2020
Guillermo Ortiz-Jimenez, Apostolos Modas, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard

* Preprint (Under review) - Accepting feedback. 23 pages, 14 figures 

  Access Paper or Ask Questions

Neural Anisotropy Directions


Jun 17, 2020
Guillermo Ortiz-Jimenez, Apostolos Modas, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard

* 39 pages, 22 figures 

  Access Paper or Ask Questions

GeoDA: a geometric framework for black-box adversarial attacks


Mar 13, 2020
Ali Rahmati, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard, Huaiyu Dai

* In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020 

  Access Paper or Ask Questions

Hold me tight! Influence of discriminative features on deep network boundaries


Feb 15, 2020
Guillermo Ortiz-Jimenez, Apostolos Modas, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard


  Access Paper or Ask Questions

A geometry-inspired decision-based attack


Mar 26, 2019
Yujia Liu, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard


  Access Paper or Ask Questions

Robustness via curvature regularization, and vice versa


Nov 23, 2018
Seyed-Mohsen Moosavi-Dezfooli, Alhussein Fawzi, Jonathan Uesato, Pascal Frossard


  Access Paper or Ask Questions

SparseFool: a few pixels make a big difference


Nov 18, 2018
Apostolos Modas, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard


  Access Paper or Ask Questions

Divide, Denoise, and Defend against Adversarial Attacks


Feb 19, 2018
Seyed-Mohsen Moosavi-Dezfooli, Ashish Shrivastava, Oncel Tuzel

* This paper has been submitted for publication on February 9, 2018 

  Access Paper or Ask Questions

Adaptive Quantization for Deep Neural Network


Dec 04, 2017
Yiren Zhou, Seyed-Mohsen Moosavi-Dezfooli, Ngai-Man Cheung, Pascal Frossard

* 9 pages main paper + 5 pages supplementary, 8 figures, conference 

  Access Paper or Ask Questions

Geometric robustness of deep networks: analysis and improvement


Nov 24, 2017
Can Kanbak, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard


  Access Paper or Ask Questions

Analysis of universal adversarial perturbations


May 26, 2017
Seyed-Mohsen Moosavi-Dezfooli, Alhussein Fawzi, Omar Fawzi, Pascal Frossard, Stefano Soatto


  Access Paper or Ask Questions

Classification regions of deep neural networks


May 26, 2017
Alhussein Fawzi, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard, Stefano Soatto


  Access Paper or Ask Questions

Universal adversarial perturbations


Mar 09, 2017
Seyed-Mohsen Moosavi-Dezfooli, Alhussein Fawzi, Omar Fawzi, Pascal Frossard

* Accepted at IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017 

  Access Paper or Ask Questions

Robustness of classifiers: from adversarial to random noise


Aug 31, 2016
Alhussein Fawzi, Seyed-Mohsen Moosavi-Dezfooli, Pascal Frossard

* Accepted to NIPS 2016 

  Access Paper or Ask Questions

DeepFool: a simple and accurate method to fool deep neural networks


Jul 04, 2016
Seyed-Mohsen Moosavi-Dezfooli, Alhussein Fawzi, Pascal Frossard

* In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016 

  Access Paper or Ask Questions