TriDetect: We propose TriDetect tool that secures the digital ecosystem. TriDetect provides not just GenAI detection, but actionable source attribution and evidentiary explanation.
Data Free Model Extraction (DFME): This project aims to evaluate the resilience of image classification models against DFME in Machine Learning As-A-Service (MLaaS) settings through developing novel DFME techniques to improve extraction accuracy and query efficiency through advanced attack strategies to push the limits of existing DFME methods.
Deepfake detection: The project aims to developing new Deep Learning-based methods that can enhance the robustness and generalization of passive detectors and the robustness of proactive detectors.
UAV Security and Forensics: The project aims to deliver a validated, privacy-preserving, and secure SDN-enabled AI-driven UAV framework for different application domains (collaboration with UCD CCI Centre, EU ECTEG group).
LLM Safety and Integrity: Manipulating Prompts and Retrieval-Augmented Generation for LLM Service Providers
EU ENFIELD: Generalization of Deepfake Voice Detection across Generator Architectures
SDN-IoT: SDN-IoT intrusion detection dataset generated from a hybrid testbed integrating physical IoT devices with SDN control-plane telemetry. The dataset is designed to support reproducible, cross-layer security research in next-generation programmable IoT networks.
AI for healthcare: Research and development of advanced deep learning models for cardiovascular disease prediction (collaboration with National University of Vietnam at Ho Chi Minh City, and Vietnamese Hospitals)
FraudLens: An AI-based platform to aid law enforcement and financial institutions in end-to-end investigation of cryptocurrency related crime
CloudAtlas: System and methods for Verification and Auditing of Intelligent systems
CERBERUS: Child Exploitation Response by Beating Encryption and Research to Unprotect Systems
Blood Brother BB/1.0: A digital health platform that empowers people to better understand and hence manage their risk of developing diabetes.
FormIN: Forensic Readiness Methodology for IoT Network Investigation
SILAS: Securing Machine Learning Models in Cyber Security – Adversarial Science based Model and Approach
IoT Security and Forensics
UrbanArk: (collaborate with Prof. M. Bertolotto)
CONSUS: (collaborate with Prof. T. Kechadi)
InSDN (collaborate with Dr. A. Jurcut): A secure framework to automatically detect and prevent attacks in software defined network (SDN) based cloud computing environment
>> SDN Dataset
Electromagnetic Side-Channel (EM-SCA) Analysis for Security and Digital Forensics: Framework and Datasets (DOI: https://doi.org/10.1109/ACCESS.2021.3104525)
Blockchain and Cryptocurrencies:
Smart Vehicle Security and Forensics
WhatsApp Interception
CyberForensicsONT
AI Model for Efficiently Reviewing and Analysing Prospectus Documents