• Artificial Intelligence
• Environmental Monitoring
• Geospatial Analysis
• Earth Observation
• Earth Imagery Processing
• Time-series Image Analysis
• Change Detection
• Bushfire Risk Analysis
• Risk Modelling and Reduction
• Natural Hazards
• GIS Maps Updating
• Vegetation Dynamics
• Land Cover Mapping
• Advanced Machine Learning
• Explainable AI (XAI)

Bushfire Data Challenges Program under the Australian Research Data Common (ARDC)

• The goal of this project is to create aggregated and harmonized datasets for fuel attributes known to influence the fire behavior models on a national scale (Australia) using Earth observation techniques and Artificial Intelligence (AI) to solve the current data challenges in bushfire research with the aim of improving Australia’s bushfire resilience, response, and recovery. The national fuel attributes databases this project will produce will have direct benefits to bushfire planning and response given the fuel data will also be readably available for assessing bushfire risk, predicting fire behaviour, informing suppression efforts and planning prescribed burns.

Forecasting Grass Pollen with Satellite Sensor Time-series, Meteorology Data, and Machine Learning Tools

• In this project, which was a cross-faculty project between the Faculty of Engineering and Science and Health Department at the University of Technology Sydney (UTS), a consistent set of satellite measures of grass cover and seasonal greenness were investigated to evaluate their utility for predicting airborne grass pollen concentrations based on Artificial Intelligence methods.

Land Cover Classification and GIS Maps Updating from Remote Sensing Data Using Deep Convolutional Neural Networks (DCNNs)

• Land cover is the detected bio-physical overlay on the Earth’s surface, including materials like grass, forest, pastures, built-up areas, water body, etc. Reliable information on global land cover is required to assist in the solution of a wide range of environmental problems. The use of deep learning (DL) approaches for the analysis of remote sensing (RS) data is rapidly increasing. DL techniques have provided excellent results in a range of applications, especially in image processing, land cover classification, and maps updating. Thus, this work aims to deepen the understanding of DCNN models for land use/land cover classification based on time series satellite images in the context of large-scale areas.