Ongoing works
- Efficient urban traffic data processing using Federated learning
- Objectives
- To design and implement the efficient method for large-scaed urban traffic data processing using federated learning
- To develop the optimal training scheduling and organizations of local models in urban scale
- To develop the parameter sharing algorithm to converge the given multiple models
- Objectives
Previous works
- Vision-based road safety and risk analysis framework
- Objectives
- To develop a potential pedestrian risk stiatuon analysis system
- To develop a vision-based feature extraction algorithm in automatic
- Objectives
- Development of a cooperative-automatic driving simulator and demand response path calculation algorithm
- Objectives
- To develop a pedestrian trajectory prediction model in CCTV-vision environment
- To implement a living lab for management and operation of cooperative-automatic driving bus based on predictive information about road status (e.g. pedestrian trajectory) in Sejong City
- To design scenario-based system architecture for C-ITS management and operation
- Objectives
- Seeing the city: Flows, space, and data
- Objectives
- To survey the trend of pedestrian eath by road environment in urban
- To analyze the correlation bettwen spped camera deployment and traffic collisions
- Objectives
- Interdisciplinary research on smart health care based on residents’ behavior analysis
- Objectives
- To analyze the behavior patterns of the elderly in the room, and collect their health information by using statistical methods
- Objectives