Dr A M Vincent

Profile

I am a Postdoctoral Research Assistant in the School of Geography at the University of Leeds. My research lies at the intersection of artificial intelligence, remote sensing and environmental hazard assessment, with a particular focus on developing AI-based approaches for monitoring environmental change from satellite imagery.

I currently work on glacial lake monitoring and flood hazard assessment as part of the Glacial Lake Observatory (GLO). My current research focuses on developing automated deep-learning workflows using Sentinel-1 Synthetic Aperture Radar (SAR) imagery to detect and map glacial lakes across High Mountain Asia. This work aims to improve large-scale monitoring of rapidly changing glacial lakes and support the assessment of glacial lake outburst flood (GLOF) hazards.

Prior to joining the University of Leeds, I was a Postdoctoral Researcher at TU Delft, where I worked on the EU Horizon Europe SeaClear project. My research focused on artificial intelligence for underwater perception and marine litter detection, including real-time object detection and the generation of synthetic underwater imagery using generative AI.

I completed my PhD in Machine Learning at the National Institute of Technology Karnataka, India. My doctoral research focused on machine learning and deep-learning methods for geospatial applications, including flood susceptibility mapping, land-use and land-cover classification, satellite image analysis and hyperparameter optimisation.

Research interests

My research interests lie at the intersection of artificial intelligence, Earth observation and environmental science. I am particularly interested in how machine learning and deep learning can be used with satellite and geospatial data to observe environmental change and better understand natural hazards.

A key focus of my research is Geospatial AI (GeoAI), including computer vision methods for extracting meaningful information from complex Earth observation data. I am interested in image segmentation and classification, object detection, multi-temporal analysis, and the use of Synthetic Aperture Radar (SAR) and other remote-sensing datasets for environmental applications.

I am also interested in developing robust and computationally efficient AI methods for situations where environmental observations are sparse, noisy or difficult to acquire. This includes generative AI and synthetic data, model optimisation, and approaches that improve the generalisability of machine-learning models across different environmental conditions and geographical regions.

More broadly, my research aims to bridge methodological advances in AI with practical challenges in environmental monitoring, climate resilience and disaster-risk assessment.

Qualifications

• PhD in Machine Learning, National Institute of Technology Karnataka, India (2020–2024)
• MTech in Computational Mathematics, National Institute of Technology Karnataka, India (2017–2019)
• BTech in Computer Science, University of Calicut, India (2012–2016)