Dr Dimitrios Bellos

Dimitrios is a researcher and software developer, specialised in the development of deep learning, AI and machine learning solutions for the processing of Electron Micrograms (EM) and X-ray Tomograms (X-ray CTs). Namely, he researches deep learning solutions for the visual quality improvement of EMs and CTs without the use of external training datasets. Prior to this he worked as Post-doc for the University of Oxford, received his PhD in Computer Science from the University of Nottingham and his BSc and MSc Degree in Electrical and Computer Engineering from Aristotle University of Thessaloniki in Greece.

During his PhD, he worked jointly for the University of Nottingham and Diamond Light Source for the development of deep learning approaches regarding the denoising and semantic segmentation of undersampled micro X-ray CTs present in tomogram time-series, also referred to as 4D (3D+time) datasets.

Since his thesis submission, he also worked as a Postdoc for the University of Oxford regarding the suppression of missing wedge artifacts present in EM using deep learning solutions that do not involve external training dataset.

His main research interests are regarding the development of AI solutions with focus on denoising and segmentation, deep/machine learning, computer vision and image processing.