Publications associées à Procédure accusatoire et contradictoire

Understanding generalization and robustness in modern deep learning

In this thesis, we study two closely related directions: robustness and generalization in modern deep learning. Deep learning models based on empirical risk minimization are known to be often non-robust to small, worst-case perturbations known as adversari ...

EPFL2024

Natural-Looking Adversarial Examples From Freehand Sketches

Davide Nanni, Hakgu Kim

Deep neural networks (DNNs) have achieved great success in image classification and recognition compared to previous methods. However, recent works have reported that DNNs are very vulnerable to adversarial examples that are intentionally generated to misl ...

IEEE2022

Towards Robust Vision Transformer

Shaokai Ye, Yuan He

Recent advances on Vision Transformer (ViT) and its improved variants have shown that self-attention-based networks surpass traditional Convolutional Neural Networks (CNNs) in most vision tasks. However, existing ViTs focus on the standard accuracy and com ...

IEEE COMPUTER SOC2022

Block-Sparse Adversarial Attack To Fool Transformer-Based Text Classifiers

Pascal Frossard, Seyedeh Sahar Sadrizadeh

Recently, it has been shown that, in spite of the significant performance of deep neural networks in different fields, those are vulnerable to adversarial examples. In this paper, we propose a gradient-based adversarial attack against transformer-based tex ...

IEEE2022

QAIR: Practical Query-efficient Black-Box Attacks for Image Retrieval

Shaokai Ye, Yuan He, Hang Su, Jinfeng Li

We study the query-based attack against image retrieval to evaluate its robustness against adversarial examples under the black-box setting, where the adversary only has query access to the top-1 ranked unlabeled images from the database. Compared with que ...

IEEE COMPUTER SOC2021

Robustness via Cross-Domain Ensembles

Amir Roshan Zamir, Oguzhan Fatih Kar, Shuqing Teresa Yeo

We present a method for making neural network predictions robust to shifts from the training data distribution. The proposed method is based on making predictions via a diverse set of cues (called 'middle domains') and ensembling them into one strong predi ...

IEEE2021

Mixed Nash Equilibria in the Adversarial Examples Game

Rafaël Benjamin Pinot

This paper tackles the problem of adversarial examples from a game theoretic point of view. We study the open question of the existence of mixed Nash equilibria in the zero-sum game formed by the attacker and the classifier. While previous works usually al ...

JMLR-JOURNAL MACHINE LEARNING RESEARCH2021

Watermarking-based Defense against Adversarial Attacks on Deep Neural Networks

James Richard Larus, Xiaoting Li, Dinghao Wu, Junjun Zhang

The vulnerability of deep neural networks to adversarial attacks has posed significant threats to real-world applications, especially security-critical ones. Given a well-trained model, slight modifications to the input samples can cause drastic changes in ...

IEEE2021

Fast and Provable ADMM for Learning with Generative Priors

Volkan Cevher, Armin Eftekhari, Fabian Ricardo Latorre Gomez

In this work, we propose a (linearized) Alternating Direction Method-of-Multipliers (ADMM) algorithm for minimizing a convex function subject to a nonconvex constraint. We focus on the special case where such constraint arises from the specification that a ...

2019

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