About

I am a Postdoctoral researcher in Deep Learning and Spatial Ecology, working with Prof Ben Sheldon at the University of Oxford. My research focuses on developing computational Deep Learning methods to infer species identity and phenology at the level of individual trees from time series of drone images. The aim of my work is to leverage Deep Learning, Computer Vision, and Remote Sensing technologies to enhance our understanding of ecosystems and biodiversity.

Prior to that I was an EPSRC funded PhD candidate, working on Deep Learning for Remote Sensing systems at the University of Sussex, advised by Prof. Andy Philippides and Prof. Novi Quadrianto. My research focused on extracting salient feature representations for domain adaptation and semantic segmentation tasks.

Being part of the Predictive Analytics Lab (PAL) – an interdisciplinary research team, I was involved in a British Academy funded project focused on Satellite/Aerial Image Scene Segmentation, where I designed novel Deep Learning models that classify land use from satellite/aerial images. For more details please see: - PAL. The aim of the project was to apply deep learning techniques, to map peri-urban agriculture in Ghaziabad India, and research ways of integrating multiple types of data through a web-based mapping and visualisation tool. This project contributed to the SDG 11: Sustainable Cities and Communities. Please see the demo below. demo Currently, I am developing a novel architecture that extracts salient features whilst capturing more context and spatial image information by employing a wider receptive field, thus enhancing semantic segmentation performance in aerial images. The proposed approach is efficient in detecting water bodies (lakes, rivers) from Unmanned Aerial Vehicles (UAVs). Compared to the current state of the art, the proposed model is resilient to shadows and effectively retrieves scene information that is hidden beneath canopy. This project contributes to the SDG 6: Clean Water and Sanitation.

Furthermore, I investigated the Deep Learning robustness to domain shifts due to seasonal variations and proposed a novel architecture that extracts invariant feature representations between domains by combining salient texture feature extraction and a wider receptive field. This makes the model robust in the presence of domain shifts due to seasonal variations. This project contributes to the SDG 13: Climate Action.

During my research internship at Satellite Applications Catapult (under the supervision of Dr Cristian Rossi), I applied physics aware AI and Remote Sensing data to detect and classify cement plants in China by exploiting physical properties such as plant surrounding temperature and soil moisture. This project contributes to the SDG 9: Industry, Innovation and Infrastructure.

Prior my PhD, I worked as a Control and Automation Design Engineer for the oil and gas industry, where I was exposed to the full software and hardware life cycle, systems sensing and control, communications, functional safety, project planning and management.

I am always up for new collaborations, drop me an email if you want to chat!

Dr Georgios Voulgaris

Recent News