
Explores the connection between statistical learning theory and encrypted database attacks, presenting novel approaches for database reconstruction from access pattern leakage with improved efficiency and broader applicability.

Innovative approach to automatic program repair using safety properties for precise and complete vulnerability fixes, demonstrated on 32 real-world vulnerabilities across diverse applications.

Novel attack techniques targeting NFC-enabled smartphone touchscreens, combining device fingerprinting and ghost touch generation to manipulate user interactions and potentially gain remote control.

Explore acoustic side channels in screens, revealing how subtle noises can expose displayed content and user inputs, posing privacy risks in various scenarios.

Innovative fuzzing technique using neural networks to approximate program behavior, enabling more efficient bug discovery and higher code coverage in software testing.

Efficient techniques for analyzing transformation-based password cracking software, enabling improved attack configurations and scientific reasoning about real-world password-guessing attacks.

Explores a novel approach to automatic program repair using safety properties to generate precise and comprehensive vulnerability patches, avoiding limitations of example-based methods.

Explores misconceptions about HTTPS security among end users and administrators, revealing gaps in understanding and providing recommendations for improved protocol design and user interfaces.

Exploring Certificate Transparency adoption, challenges, and impact on web security. Analyzes deployment strategies, error rates, and lessons for future large-scale security implementations.

Unveiling RIDL: a new class of speculative execution attacks that leak arbitrary data across boundaries, bypassing existing mitigations and posing significant security risks for modern processors.

Investigates ECC memory's vulnerability to Rowhammer attacks, revealing limitations in its protective capabilities through reverse engineering and a novel attack method called ECCploit.

Analyzing economic models of offline password cracking to predict attacker behavior and evaluate password protection methods, emphasizing the importance of memory-hard functions for enhanced security.

Explores predictability of biometric features across contexts, revealing vulnerabilities in behavioral biometrics and assessing resilience against cross-context attacks through a comprehensive user study.

Exploring security and privacy challenges in multi-user augmented reality through a qualitative study with HoloLens, revealing unique concerns and design implications for immersive AR technologies.

Explore targeted audio adversarial examples on automatic speech recognition, constructing waveforms that transcribe as chosen phrases while remaining nearly identical to the original.

Innovative approach to automatically construct CPS models using supervised machine learning on data from systematically seeded software faults, demonstrating efficacy in detecting various attacks on a water purification plant simulator.

Explore practical foundations for software Spectre defenses, covering key strategies and techniques to mitigate this critical security vulnerability in modern processors.

Explore efficient and scalable path constraints fuzzing techniques with Jigsaw, enhancing software testing and vulnerability detection for improved system security and reliability.

Explore how service worker caches can amplify web attacks, examining vulnerabilities and potential security risks in modern web applications.

Explore vulnerabilities in Zeroconf protocols, focusing on man-in-the-middle attacks. Learn about potential security risks and implications for network configurations.

Explore a hardware-assisted approach for unpacking Android apps, enhancing security analysis and malware detection through innovative techniques and methodologies.