Publications by Seyed Mohsen Moosavi Dezfooli

Certified Human Trajectory Prediction

Alexandre Massoud Alahi, Seyed Mohsen Moosavi Dezfooli, Saeed Saadatnejad, Mohammadhossein Bahari

Predicting human trajectories is essential for the safe operation of autonomous vehicles, yet current data-driven models often lack robustness in case of noisy inputs such as adversarial examples or imperfect observations. Although some trajectory predicti ...

2025

Sparse Attacks for Manipulating Explanations in Deep Neural Network Models

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Ahmad Ajalloeian, Michail Vlachos

We investigate methods for manipulating classifier explanations while keeping the predictions unchanged. Our focus is on using a sparse attack, which seeks to alter only a minimal number of input features. We present an efficient and novel algorithm for co ...

Ieee Computer Soc2023

On Smoothed Explanations: Quality and Robustness

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Ahmad Ajalloeian, Michail Vlachos

Explanation methods highlight the importance of the input features in taking a predictive decision, and represent a solution to increase the transparency and trustworthiness in machine learning and deep neural networks (DNNs). However, explanation methods ...

Assoc Computing Machinery2022

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

Alexandre Massoud Alahi, Seyed Mohsen Moosavi Dezfooli, Saeed Saadatnejad, Mohammadhossein Bahari, Pedram Khorsandi

Our transportation field has recently witnessed an arms race of neural network-based trajectory predictors. While these predictors are at the core of many applications such as autonomous navigation or pedestrian flow simulations, their adversarial robustne ...

2022

Vehicle trajectory prediction works, but not everywhere

Alexandre Massoud Alahi, Seyed Mohsen Moosavi Dezfooli, Saeed Saadatnejad, Mohammadhossein Bahari, Ahmad Rahimi

Vehicle trajectory prediction is nowadays a fundamental pillar of self-driving cars. Both the industry and research communities have acknowledged the need for such a pillar by providing public benchmarks. While state-of-the-art methods are impressive, i.e. ...

IEEE2022

PRIME: A Few Primitives Can Boost Robustness to Common Corruptions

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas, Guillermo Ortiz Jimenez

Despite their impressive performance on image classification tasks, deep networks have a hard time generalizing to unforeseen corruptions of their data. To fix this vulnerability, prior works have built complex data augmentation strategies, combining multi ...

SPRINGER INTERNATIONAL PUBLISHING AG2022

What can linearized neural networks actually say about generalization?

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Guillermo Ortiz Jimenez

For certain infinitely-wide neural networks, the neural tangent kernel (NTK) theory fully characterizes generalization, but for the networks used in practice, the empirical NTK only provides a rough first-order approximation. Still, a growing body of work ...

2021

Neural Anisotropy Directions

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas, Guillermo Ortiz Jimenez

In this work, we analyze the role of the network architecture in shaping the inductive bias of deep classifiers. To that end, we start by focusing on a very simple problem, i.e., classifying a class of linearly separable distributions, and show that, depen ...

2020

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

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas, Guillermo Ortiz Jimenez

Driven by massive amounts of data and important advances in computational resources, new deep learning systems have achieved outstanding results in a large spectrum of applications. Nevertheless, our current theoretical understanding on the mathematical fo ...

2020

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

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas, Guillermo Ortiz Jimenez

Important insights towards the explainability of neural networks reside in the characteristics of their decision boundaries. In this work, we borrow tools from the field of adversarial robustness, and propose a new perspective that relates dataset features ...

2020

Redundant features can hurt robustness to distribution shift

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas, Guillermo Ortiz Jimenez

In this work, we borrow tools from the field of adversarial robustness, and propose a new framework that permits to relate dataset features to the distance of samples to the decision boundary. Using this framework we identify the subspace of features used ...

2020

Robustness via curvature regularization, and vice versa

Pascal Frossard, Alhussein Fawzi, Seyed Mohsen Moosavi Dezfooli

State-of-the-art classifiers have been shown to be largely vulnerable to adversarial perturbations. One of the most effective strategies to improve robustness is adversarial training. In this paper, we investigate the effect of adversarial training on the ...

IEEE2019

Universal Adversarial Attacks on Text Classifiers

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Melika Behjati

Despite the vast success neural networks have achieved in different application domains, they have been proven to be vulnerable to adversarial perturbations (small changes in the input), which lead them to produce the wrong output. In this paper, we propos ...

IEEE2019

A Geometry-Inspired Decision-Based Attack

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Yujia Liu

Deep neural networks have recently achieved tremendous success in image classification. Recent studies have however shown that they are easily misled into incorrect classification decisions by adversarial examples. Adversaries can even craft attacks by que ...

IEEE COMPUTER SOC2019

SparseFool: a few pixels make a big difference

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas

Deep Neural Networks have achieved extraordinary results on image classification tasks, but have been shown to be vulnerable to attacks with carefully crafted perturbations of the input data. Although most attacks usually change values of many image's pixe ...

IEEE2019

Geometry of adversarial robustness of deep networks: methods and applications

Seyed Mohsen Moosavi Dezfooli

We are witnessing a rise in the popularity of using artificial neural networks in many fields of science and technology. Deep neural networks in particular have shown impressive classification performance on a number of challenging benchmarks, generally in ...

EPFL2019

Adaptive Quantization for Deep Neural Network

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Yiren Zhou

In recent years Deep Neural Networks (DNNs) have been rapidly developed in various applications, together with increasingly complex architectures. The performance gain of these DNNs generally comes with high computational costs and large memory consumption ...

ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE2018

SparseFool: a few pixels make a big difference

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Apostolos Modas

Deep Neural Networks have achieved extraordinary results on image classification tasks, but have been shown to be vulnerable to attacks with carefully crafted perturbations of the input data. Although most attacks usually change values of many image's pixe ...

2018

Robustness of Classifiers to Universal Perturbations: A Geometric Perspective

Pascal Frossard, Alhussein Fawzi, Seyed Mohsen Moosavi Dezfooli

Deep networks have recently been shown to be vulnerable to universal perturbations: there exist very small image-agnostic perturbations that cause most natural images to be misclassified by such classifiers. In this paper, we provide a quantitative analysi ...

2018

Geometric robustness of deep networks: analysis and improvement

Pascal Frossard, Seyed Mohsen Moosavi Dezfooli, Can Kanbak

Deep convolutional neural networks have been shown to be vulnerable to arbitrary geometric transformations. However, there is no systematic method to measure the invariance properties of deep networks to such transformations. We propose ManiFool as a simpl ...

2018