Briland Hitaj

Advanced Computer Scientist II

Computer Science Laboratory, SRI International

briland [dot] hitaj [at] sri [dot] com

hitaj [at] di [dot] uniroma1 [dot] it

Google Scholar - DBLP

I am an Advanced Computer Scientist at the Computer Science Laboratory (CSL) of SRI International.

Previously, I was a Visiting Research Scholar at the Department of Computer Science at Stevens Institute of Technology working with Prof. Giuseppe Ateniese in the intersection between security, privacy and deep learning. 

On October 2018, I earned my Ph.D. in Computer Science from Department of Computer Science at Sapienza University of Rome. I consider myself lucky and honored to have been advised from and worked alongside Prof. Luigi V. Mancini and Prof. Giuseppe Ateniese.

My research interests include but are not limited to security, privacy, deep learning, generative adversarial networks (GANs) uses in security & privacy related problems, (distributed) privacy-preserving machine learning, cyber-intelligent agents, application and incorporation of deep learning in cyber-security domain.

Highlighted Publications

MaleficNet: Hiding Malware into Deep Neural Networks Using Spread-Spectrum Channel Coding

Dorjan Hitaj, Giulio Pagnotta, Briland Hitaj, Luigi V. Mancini, Fernando Perez-Cruz

In this paper, we push the boundaries of current state-of-the-art by introducing MaleficNet, a technique that combines spread-spectrum channel coding with error correction techniques, injecting malicious payloads in the parameters of deep neural networks, all while causing no degradation to the model’s performance and successfully bypassing state-of-the-art detection and removal mechanisms.

Venue: ESORICS'22  [PDF]

FedComm: Federated Learning as a Medium for Covert Communication

Dorjan Hitaj, Giulio Pagnotta, Briland Hitaj, Fernando Perez-Cruz, Luigi V. Mancini

We introduce FedComm, a novel covert-communication technique that enables robust sharing and transfer of targeted payloads within the FL framework. Our extensive theoretical and empirical evaluations show that FedComm provides a stealthy communication channel, with minimal disruptions to the training process.

Venue: IEEE TDSC [PDF]

PassGAN: A Deep Learning Approach for Password Guessing

Briland Hitaj, Paolo Gasti, Giuseppe Ateniese, Fernando Perez-Cruz

In this paper, we introduce PassGAN, a novel approach that replaces human-generated password rules with theory-grounded machine learning algorithms. PassGAN autonomously learns the distribution of real passwords from actual password leaks, and generates high-quality password guesses without any a-priori knowledge on passwords or common password structures.

Venue: ACNS'19

 SECML'18 (co-located with NeurIPS)  [PDF] - no proceedings

Deep Models Under the GAN

Deep Models Under the GAN: Information Leakage from Collaborative Deep Learning

Briland Hitaj, Giuseppe Ateniese, Fernando Perez-Cruz

We show that any privacy-preserving collaborative deep learning is susceptible to a powerful attack that we devise in this paper. The attack we developed exploits the real-time nature of the learning process that allows the adversary to train a Generative Adversarial Network (GAN) that generates prototypical samples of the targeted training set that was meant to be private.

Venue: ACM CCS'17  [PDF] [SLIDES] [PRESENTATION]

No Place to Hide that Bytes won't Reveal: Sniffing Location-Based Encrypted Traffic to Track a User's Position

Giuseppe Ateniese, Briland Hitaj, Luigi V. Mancini, Nino V. Verde, Antonio Villani

In this paper, we propose a new adversary model for Location Based Services (LBSs). The model takes into account an unauthorized third party, different from the LBS provider itself, that wants to infer the location and monitor the movements of a LBS user. We show that such an adversary can extrapolate the position of a target user by just analyzing the size and the timing of the encrypted traffic exchanged between that user and the LBS provider. 

Venue: NSS'15  [PDF] [SLIDES]

Press Coverage

News