Spatial AI and Knowledge Integration Lab

About

Our Research

Welcome to our research lab at the Department of Computer Science, National Yang Ming Chiao Tung University.

We develop spatial AI methods for integrating, interpreting, and reasoning over heterogeneous geospatial, visual, textual, and historical data. Our research focuses on methods that:

  • Represent or link entities from heterogeneous data sources.
  • Exploit learned or symbolic knowledge to interpret linked entities and their behaviors.
  • Support interpretable or optimal decision-making under multiple constraints.

We study these research questions across domains such as geology, transportation, education, infrastructure, logistics, human mobility, and historical geography.

Research themes

What We Work on

Spatial AI and Geographic Intelligence

Geographic information science, spatiotemporal reasoning, urban computing, human mobility.

Visual and Document Understanding

Polygonal entity linking, visual-language reasoning, historical map understanding, conditioned instance segmentation.

Knowledge Integration and Semantic Reasoning

Heterogeneous data integration, spatial semantics, concept-structure reasoning, knowledge-augmented modeling.

Decision Intelligence and Optimization

Multi-criteria decision analysis, constraint satisfaction, resource allocation and scheduling, learning-augmented optimization.

News

Recent Updates

Aug 2026
SAKI Lab will launch in the Department of Computer Science at NYCU.
Publications

Selected Publications

Loading selected publications...

Full Publication List
Support

Funding and Support