민덕기 연구실

Edge Computing brings computation and data storage closer to data sources. Fog Computing uses edge devices for computation, storage and communication locally. Research includes Autonomic Provisioning Edge Cluster System and Smart Gateway for IoT devices.

Cloud computing delivers computing resources as a service with rapid provisioning and minimal management effort. Research focuses on Autonomic Provisioning Cloud Computing System, Serverless computing, and Reinforcement Learning for Cloud Computing.

Research on global and local navigation for mobile robots including path planning, SLAM, obstacle avoidance, and digital twin simulations. Covers wheeled robots, UAVs, and multi-agent coordination using Deep Reinforcement Learning.

Computer vision research focused on interactive healthcare cleaning robot system. Applications include object detection, face recognition, emotion recognition, action recognition, and pose estimation for elderly care assistance.

NLP research on Large Language Models including prompt optimization, parameter-efficient fine-tuning, and LLM agents. Focus on techniques like Direct Preference Optimization, Low-Rank Adaptation, and multi-agent systems for complex NLP tasks.

Digital twin technology creates computational models that represent physical systems in real-time. DMS Group develops neural digital twin dynamic engines, control engines, control frameworks, and cloud infrastructure for unmanned systems.

Research on quantification methodologies for dependability and security metrics of computing systems including virtualized servers, data centers, SDN, Cloud-Fog-Edge continuum, IoMT, IoIT, and unmanned aerial systems.