I conduct research at the intersection of artificial intelligence, computer vision, and context-aware systems, with an emphasis on designing and optimizing efficient, reliable learning models that operate across heterogeneous computing platforms, including resource-constrained edge devices and conventional high-performance systems. My work integrates deep neural architectures with domain knowledge and contextual signals to support real-world deployments in smart agriculture, embedded forensics, and autonomous sensing, where trade-offs among accuracy, scalability, latency, and robustness are critical. 

A central theme of my recent research is model–system co-design, examining how learning frameworks can be structured to adapt to varying execution environments and resource constraints. This includes developing context-aware, multimodal learning models, edge-efficient architectures, and knowledge- and LLM-assisted approaches that improve interpretability, robustness, and decision support in scientific and cyber-physical systems.