Soheib new doctor in Computer Engineering

In recent years, the demand for immersive and efficient imaging solutions has increased with applications in areas that require both human interpretation and automated computer vision, such as remote control, augmented reality, and telepresence. Soheib's research addresses the challenges of representing and compressing multi-sensor video data, especially in the context of light-field imaging, which captures both angular and spatial information. Light-field imaging has the potential to significantly improve human perception and machine vision by offering additional depth and perspective clues that traditional 2D imaging lacks. However, current methods of light-field image compression are often computationally demanding and unsuitable for real-time or resource-limited applications. Soheib's research proposes a novel framework based on variational autoencoder (VAE) for light-field image compression, with the aim of effectively distinguishing spatial and angular information to facilitate efficient data representation and compression.

The dissertation took place at Campus in Sundsvall. Supervisor and chairman of the dissertation was Professor Mårten Sjöström from Mid Sweden University and Head of Research Christine Guillemot, INRIA Rennes, France, and  the opponent was Professor João Ascenso, Instituto Superior Técnico, Lisbon, Portugal. The examining committee consisted of Associate Professor Markus Flierl, KTH Royal Institute of Technology, Professor Carl Debono, University of Malta and Associate Professor Giuseppe Valenzise, Université Paris-Saclay, France.

Read the thesis in Diva