My research contributions span applied machine learning, signal processing, and inverse modeling across multiple domains. While my doctoral research focuses on reconfigurable wave-based systems for next-generation communications, my broader work develops general methodologies for inverse, constrained, and physics-consistent learning, with applications in seismic analysis, imaging, healthcare, sensing, and data-driven system modeling.
This paper pioneers a first-of-its-kind Bidirectional Liquid Neural Network (BLiqNet) framework to generate highly accurate, physically-informed synthetic seismic data. This breakthrough significantly reduces engineering errors to advance the state of the art in earthquake hazard modeling.
We employ a Conditional Generative Adversarial Network (CGAN) to accurately model and predict the acceleration response spectra, critical for structural safety analysis in earthquake engineering.
Developed SLICE, a novel image processing methodology combining clustering and Earth Mover’s distance, for highly accurate and robust classification of brain tumours in medical imagery.
A detailed review exploring the synergy between photoacoustic imaging techniques and functional nanoparticles for enhanced biomedical diagnostics and therapeutic applications.
This review synthesizes the current state of nanotechnology in creating high-performance near-infrared photodetectors, focusing on material science and performance metrics.
A quality engineering study analyzing the Critical-to-Quality (CTQ) factors for Uninterruptable Power Supply (UPS) systems to improve reliability and performance standards.
Patent Application No.: 202341052891: A novel system design focused on real-time multi-object tracking and classification in complex environments, integrating sensor fusion.
Patent Application No.: 202341075746: Proposes an integrated, secure IoT framework specifically designed for continuous health monitoring and data analytics in clinical and home settings.
Patent Application No.: 202441039763: A system capable of simulating and performing complex analyses on surface bioluminescence patterns for marine biology or environmental studies.
Patent Application No.: 202541076849: Describes an efficient hardware and software architecture for optimized frequency domain signal processing in OFDM (Orthogonal Frequency-Division Multiplexing) systems, relevant to 5G/6G.
Available to view here: Master's Thesis
Available to view here: Bachelor's Thesis