Research Overview
My research focuses on the intersection of urban geography, artificial intelligence, and geospatial data analysis. I develop innovative approaches to understand urban dynamics and improve urban resilience, livability through advanced computational methods.
Current Projects
- Urban Flash flood prediction: Predicting Urban Flash Flood combining GeoAI and multiple geospatial data sources.
- Building use and livability simultaneously prediction: For this multi-task research, exploring using multiple geospatial data and GeoAI to predict both builidng use information and its livability score at the same time.
Research Areas
Urban resilience evaluation
- Multi-modal Deep Learning for flood depth prediction and mapping: Exploring how to predict flood depth according to diverse geospatial data, including time-serious data. A state-of-the-art deep learnign method will be proposed and evaluated.
Urban spatial structure and unequality analysis
- Multi-modal Deep Learning for Urban Livability: Developing comprehensive frameworks to evaluate and measure urban quality of life using multiple data sources including remote sensing, social media, and urban infrastructure data
- Building Functional Analysis: Extracting and analyzing building use patterns from various geospatial datasets to understand urban functional diversity
- Urban Property Valuation: Using machine learning and geospatial analysis to assess property values and urban development patterns
Geospatial AI and Remote Sensing
- Deep Learning Applications: Applying multi-modal deep learning techniques to urban research problems
- High-resolution Remote Sensing: Processing and analyzing high-resolution satellite imagery for urban land use classification
- Crop Mapping: Developing automated methods for agricultural land use mapping and monitoring
Data Integration and Analysis
- Geospatial Data Integration: Combining remote sensing, GIS, and other spatial data for comprehensive urban analysis
- Mixed-Use Building Classification: Advanced methods for identifying and classifying complex urban building functions
- Spatial Statistics: Applying statistical methods to understand spatial patterns and relationships in urban environments
Reviewer
- Remote Sensing of Environment
- ISPRS Journal of Photogrammetry and Remote Sensing
- IEEE Transactions on Geoscience and Remote Sensing
- Cities
- International Journal of Remote Sensing
- International Journal of Digital Earth
- International Journal of Geographical Information Science
- IEEE Access
- Cartography and Geographic Information Science
- Asian Geography
- Computational Urban Science
Previous Research Highlights
Ph.D. Research (2018-2023)
- Estimating urban livability using multiple data sources based on multi-modal deep learning
- Developed innovative approaches to assess urban livability by integrating diverse data sources
- Applied multi-modal deep learning techniques to create comprehensive urban livability indices
- Exploring the mechanism between building spatial and functional characteristics and urban livability
- Investigated the relationships between physical building characteristics and urban quality of life
- Analyzed how building spatial patterns influence overall urban livability metrics
- Extracting hierarchical building use information from multiple data sources using deep learning methods
- Developed methods to classify buildings into broad categories, detailed categories, and mixed-use classifications
- Created hierarchical classification systems for urban building functional analysis
- Broad building functional information extraction using feature fusion based multimodal Transformer network considering mixed-use information
- Implemented advanced Transformer networks for building functional classification
- Integrated multiple data modalities to improve classification accuracy, especially for mixed-use buildings
Master’s Research (2015-2018)
- Urban functional information extraction from high-resolution remote sensing images based on deep learning
- Developed deep learning approaches for extracting urban functional information from high-resolution satellite imagery
- Implemented stratified processing methods to improve classification accuracy and efficiency