About me
I am a postdoctoral researcher in the Institue for Machine Learning at the ETH Zurich. My research interests broadly include machine learning, computer vision, and signal processing. Currently, my research focuses on the analysis of the robustness of classifiers and its connection with the geometry of deep networks. I received a PhD in Computer and Communication Sciences under the supervision of Prof. Pascal Frossard at the École Polytechnique Fédérale de Lausanne (EPFL), Switzerland. During my PhD, I interned for the Apple AI Research in 2017 where I worked on autonomous technologies. I received my M.S. degree in Communication Systems from the EPFL in 2014 where I worked as a research assistant under the supervision of Prof. Martin Vetterli. In 2012, I got my B.S. in Electrical Engineering from Amirkabir University of Technology (Tehran Polytechnic), Iran.
I am available on the academic job market.
Recent talks
A geometric perspective on the robustness of deep networks
Publications
Neural anisotropy directions
Guillermo Ortiz-Jimenez, Apostolos Modas, Seyed Moosavi, Pascal Frossard
Neural Information Processing Systems (NeurIPS), 2020.
Hold me tight! influence of discriminative features on deep network boundaries
Guillermo Ortiz-Jimenez, Apostolos Modas, Seyed Moosavi, Pascal Frossard
Neural Information Processing Systems (NeurIPS), 2020.
GeoDA: a geometric framework for black-box adversarial attacks
Ali Rahmati, Seyed Moosavi, Pascal Frossard, Huaiyu Dai
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020.
A geometry-inspired decision-based attack
Yujia Liu, Seyed Moosavi, Pascal Frossard
International Conference on Computer Vision (ICCV), 2019.
Robustness via curvature regularization, and vice versa
Seyed Moosavi, Alhussein Fawzi, Jonathan Uesato, Pascal Frossard
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019.
SparseFool: a few pixels make a big difference
Apostolos Modas, Seyed Moosavi, Pascal Frossard
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019.
Divide, Denoise, and Defend against Adversarial Attacks
Seyed Moosavi, Ashish Shrivastava, Oncel Tuzel
arXiv pre-print
Geometric robustness of deep networks
Can Kanbak, Seyed Moosavi, Pascal Frossard
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018.
Empirical study of the topology and geometry of deep networks
Alhussein Fawzi, Seyed Moosavi, Pascal Frossard, Stefano Soatto
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018.
Robustness of Classifiers to Universal Perturbations: A Geometric Perspective
Seyed Moosavi, Alhussein Fawzi, Omar Fawzi, Pascal Frossard, Stefano Soatto
International Conference on Learning Representations (ICLR), 2018.
A Geometric Perspective on the Robustness of Deep Networks
Alhussein Fawzi, Seyed Moosavi, Pascal Frossard
IEEE Signal Processing Magazine, 2017.
Universal adversarial perturbations
Seyed Moosavi*, Alhussein Fawzi*, Pascal Frossard
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
Robustness of classifiers: from adversarial to random noise
Alhussein Fawzi*, Seyed Moosavi*, Pascal Frossard
Neural Information Processing Systems (NIPS), 2016.
DeepFool: a simple and accurate method to fool deep neural networks
Seyed Moosavi, Alhussein Fawzi, Pascal Frossard
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.