148 citations · 439 across the 31 of their papers we have counts for
24 papers · 1 filter
Adversarial Frontiers: Minimum-Norm Attack Ensembles for Robustness Evaluation
Luca Scionis, Luca Melis, Maura Pintor +5
Adversarial robustness is commonly evaluated with predefined attack ensembles, such as AutoAttack, at a single perturbation budget and on a selective choice of pertur…
Out-of-Distribution Detection for Continual Learning: Design Principles and Benchmarking
Srishti Gupta, Riccardo Balia, Daniele Angioni +7
Recent years have witnessed significant progress in the development of machine learning models across a wide range of fields, fueled by increased computational resources, large-sca…
Buffer-free Class-Incremental Learning with Out-of-Distribution Detection
Srishti Gupta, Daniele Angioni, Maura Pintor +4
Class-incremental learning (CIL) poses significant challenges in open-world scenarios, where models must not only learn new classes over time without forgetting previous ones but a…
Adversarial Pruning: A Survey and Benchmark of Pruning Methods for Adversarial Robustness
Giorgio Piras, Maura Pintor, Ambra Demontis +3
Recent work has proposed neural network pruning techniques to reduce the size of a network while preserving robustness against adversarial examples, i.e., well-crafted inputs induc…
HO-FMN: Hyperparameter Optimization for Fast Minimum-Norm Attacks
Raffaele Mura, Giuseppe Floris, Luca Scionis +6
Gradient-based attacks are a primary tool to evaluate robustness of machine-learning models. However, many attacks tend to provide overly-optimistic evaluations as they use fixed l…
Over-parameterization and Adversarial Robustness in Neural Networks: An Overview and Empirical Analysis
Srishti Gupta, Zhang Chen, Luca Demetrio +9
Thanks to their extensive capacity, over-parameterized neural networks exhibit superior predictive capabilities and generalization. However, having a large parameter space is consi…
AttackBench: Evaluating Gradient-based Attacks for Adversarial Examples
Antonio Emanuele Cinà, Jérôme Rony, Maura Pintor +5
Adversarial examples are typically optimized with gradient-based attacks. While novel attacks are continuously proposed, each is shown to outperform its predecessors using differen…
Improving Fast Minimum-Norm Attacks with Hyperparameter Optimization
Giuseppe Floris, Raffaele Mura, Luca Scionis +4
Evaluating the adversarial robustness of machine learning models using gradient-based attacks is challenging. In this work, we show that hyperparameter optimization can improve fas…
Samples on Thin Ice: Re-Evaluating Adversarial Pruning of Neural Networks
Giorgio Piras, Maura Pintor, Ambra Demontis +1
Neural network pruning has shown to be an effective technique for reducing the network size, trading desirable properties like generalization and robustness to adversarial attacks…
Minimizing Energy Consumption of Deep Learning Models by Energy-Aware Training
Dario Lazzaro, Antonio Emanuele Cinà, Maura Pintor +4
Deep learning models undergo a significant increase in the number of parameters they possess, leading to the execution of a larger number of operations during inference. This expan…
Security of Deep Reinforcement Learning for Autonomous Driving: A Survey
Ambra Demontis, Srishti Gupta, Maura Pintor +6
Reinforcement learning (RL) enables agents to learn optimal behaviors through interaction with their environment and has been increasingly deployed in safety-critical applications,…
Wild Patterns Reloaded: A Survey of Machine Learning Security against Training Data Poisoning
Antonio Emanuele Cinà, Kathrin Grosse, Ambra Demontis +7
The success of machine learning is fueled by the increasing availability of computing power and large training datasets. The training data is used to learn new models or update exi…
Why Adversarial Reprogramming Works, When It Fails, and How to Tell the Difference
Yang Zheng, Xiaoyi Feng, Zhaoqiang Xia +5
Adversarial reprogramming allows repurposing a machine-learning model to perform a different task. For example, a model trained to recognize animals can be reprogrammed to recogniz…
Indicators of Attack Failure: Debugging and Improving Optimization of Adversarial Examples
Maura Pintor, Luca Demetrio, Angelo Sotgiu +4
Evaluating robustness of machine-learning models to adversarial examples is a challenging problem. Many defenses have been shown to provide a false sense of robustness by causing g…
Backdoor Learning Curves: Explaining Backdoor Poisoning Beyond Influence Functions
Antonio Emanuele Cinà, Kathrin Grosse, Sebastiano Vascon +4
Backdoor attacks inject poisoning samples during training, with the goal of forcing a machine learning model to output an attacker-chosen class when presented a specific trigger at…
BAARD: Blocking Adversarial Examples by Testing for Applicability, Reliability and Decidability
Xinglong Chang, Katharina Dost, Kaiqi Zhao +4
Adversarial defenses protect machine learning models from adversarial attacks, but are often tailored to one type of model or attack. The lack of information on unknown potential a…
The Hammer and the Nut: Is Bilevel Optimization Really Needed to Poison Linear Classifiers?
Antonio Emanuele Cinà, Sebastiano Vascon, Ambra Demontis +3
One of the most concerning threats for modern AI systems is data poisoning, where the attacker injects maliciously crafted training data to corrupt the system's behavior at test ti…
Domain Knowledge Alleviates Adversarial Attacks in Multi-Label Classifiers
Stefano Melacci, Gabriele Ciravegna, Angelo Sotgiu +4
Adversarial attacks on machine learning-based classifiers, along with defense mechanisms, have been widely studied in the context of single-label classification problems. In this p…
Do Gradient-based Explanations Tell Anything About Adversarial Robustness to Android Malware?
Marco Melis, Michele Scalas, Ambra Demontis +4
While machine-learning algorithms have demonstrated a strong ability in detecting Android malware, they can be evaded by sparse evasion attacks crafted by injecting a small set of…
secml: A Python Library for Secure and Explainable Machine Learning
Maura Pintor, Luca Demetrio, Angelo Sotgiu +3
We present \texttt{secml}, an open-source Python library for secure and explainable machine learning. It implements the most popular attacks against machine learning, including tes…
Why Do Adversarial Attacks Transfer? Explaining Transferability of Evasion and Poisoning Attacks
Ambra Demontis, Marco Melis, Maura Pintor +5
Transferability captures the ability of an attack against a machine-learning model to be effective against a different, potentially unknown, model. Empirical evidence for transfera…
On Security and Sparsity of Linear Classifiers for Adversarial Settings
Ambra Demontis, Paolo Russu, Battista Biggio +2
Machine-learning techniques are widely used in security-related applications, like spam and malware detection. However, in such settings, they have been shown to be vulnerable to a…
Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization
Luis Muñoz-González, Battista Biggio, Ambra Demontis +4
A number of online services nowadays rely upon machine learning to extract valuable information from data collected in the wild. This exposes learning algorithms to the threat of d…
Is Deep Learning Safe for Robot Vision? Adversarial Examples against the iCub Humanoid
Marco Melis, Ambra Demontis, Battista Biggio +3
Deep neural networks have been widely adopted in recent years, exhibiting impressive performances in several application domains. It has however been shown that they can be fooled…