Coping with the enemy: Advances in adversary-aware signal processing

Abstract

This paper is a first attempt to provide a unified framework for studying signal processing problems where designers have to cope with the presence of an adversary, including media forensics, watermarking, adversarial machine learning, biometric spoofing, etc. We focus on the binary decision problem and discuss which strategies the adversary can use to flip the decision output at minimal cost, including blind sensitivity attacks and hill-climbing attacks. As the defender can also play smarter by considering the presence of a rational adversary, we introduce a game-theoretic approach where some advances have been recently made. We conclude by discussing some trends raised by this game-theoretic formulation.

Decision fusion with corrupted reports in multi-sensor networks: A game-theoretic approach

“…Even in this case, it is possible that malevolent users provide a fake feedback to improve or decrease the reputation of the item under inspection [3], [4]. Other examples come from the emerging field of adversarial signal processing as exemplified in [5].…”

Source distinguishability under corrupted training

“…Binary decision in an adversarial setup is one of the most recurrent problems in AdvSP, due to its importance in many application scenarios [1]. Among binary decision problems, source identification is one of the most studied subjects, since it lies at the heart of several security-oriented disciplines, like Multimedia forensics, anomaly detection, steganalysis and so on.…”

What did your adversary believeƒ Optimal Filtering and Smoothing in Counter-Adversarial Autonomous Systems

“…Before providing the details, note that α k (·) is a density over Π, where Π is the set of all beliefs from (2). In particular, if the statespace X is continuous, then Π is a function space comprising the space of density functions; if X is finite, then Π is the unit (X − 1)dimensional simplex (corresponding to X-dimensional probability vectors).…”

“…During the last decade, the importance of defense against cyberadversarial and autonomous treats has been highlighted on numerous occasions -e.g., [1,2,3]. In this paper, we consider the design of counter-adversarial autonomous (CAA) systems.…”