相关论文: Forensic Log Based Detection For Keystroke Injecti…

Input sanitization mechanisms are widely used to mitigate vulnerabilities to injection attacks such as cross-site scripting. Static analysis tools and techniques commonly used to ensure that applications utilize sanitization functions.…

Personal Identification Numbers (PIN) are widely used as authentication method for systems such as Automated Teller Machines (ATMs) and Point of Sale (PoS). Input devices (PIN pads) usually give the user a feedback sound when a key is…

Driving behaviour is one of the primary causes of road crashes and accidents, and these can be decreased by identifying and minimizing aggressive driving behaviour. This study identifies the timesteps when a driver in different…

Debugging denotes the process of detecting root causes of unexpected observable behaviors in programs, such as a program crash, an unexpected output value being produced or an assertion violation. Debugging of program errors is a difficult…

Fault attacks consist in changing the program behavior by injecting faults at run-time in order to break some expected security properties. Applications are hardened against fault attack adding countermeasures. According to the state of the…

While attack graphs are useful for identifying major cybersecurity threats affecting a system, they do not provide operational support for determining the likelihood of having a known vulnerability exploited, or that critical system nodes…

Identifying the tasks a given piece of malware was designed to perform (e.g. logging keystrokes, recording video, establishing remote access, etc.) is a difficult and time-consuming operation that is largely human-driven in practice. In…

Jailbreaks have been a central focus of research regarding the safety and reliability of large language models (LLMs), yet the mechanisms underlying these attacks remain poorly understood. While previous studies have predominantly relied on…

Backdoor attacks aim to surreptitiously insert malicious triggers into DNN models, granting unauthorized control during testing scenarios. Existing methods lack robustness against defense strategies and predominantly focus on enhancing…

Poison-only Clean-label Backdoor Attacks aim to covertly inject attacker-desired behavior into DNNs by merely poisoning the dataset without changing the labels. To effectively implant a backdoor, multiple \textbf{triggers} are proposed for…

Cybersecurity of Industrial Control Systems (ICS) is drawing significant concerns as data communication increasingly leverages wireless networks. A lot of data-driven methods were developed for detecting cyberattacks, but few are focused on…

Sensors (e.g., light, gyroscope, accelerotmeter) and sensing enabled applications on a smart device make the applications more user-friendly and efficient. However, the current permission-based sensor management systems of smart devices…

Prompt injection attacks can compromise the security and stability of critical systems, from infrastructure to large web applications. This work curates and augments a prompt injection dataset based on the HackAPrompt Playground Submissions…

A drive by download is a download that occurs without users action or knowledge. It usually triggers an exploit of vulnerability in a browser to downloads an unknown file. The malicious program in the downloaded file installs itself on the…

Intrusion Detection Systems (IDSs) are a necessary cyber defense mechanism. Unfortunately, their capability has fallen behind that of attackers. This motivates us to improve our understanding of the root causes of their false-negatives. In…

Multiple choice questions are at the heart of many standardized tests and examinations at academic institutions allover the world. In this paper, we argue that recent advancements in sensing and human-computer interaction expose these types…

The magnitude of force used in a stabbing incident can be difficult to quantify, although the estimate given by forensic pathologists is often seen as `critical' evidence in medico-legal situations. The main objective of this study is to…

Rowhammer is a security vulnerability that allows unauthorized attackers to induce errors within DRAM cells. To prevent fault injections from escalating to successful attacks, a widely accepted mitigation is implementing fault checks on…

There is growing adoption of smart devices such as digital locks with remote control and sophisticated authentication mechanisms. However, a lack of attention to device security and user-awareness beyond the primary function of these IoT…

Detecting kinetic vulnerabilities in Cyber-Physical Systems (CPS), vulnerabilities in control code that can precipitate hazardous physical consequences, is a critical challenge. This task is complicated by the need to analyze the intricate…