• Large-Scale Detection and Categorization of Oil Spills from SAR Images with Deep Learning 

      Bianchi, Filippo Maria; Espeseth, Martine; Borch, Njål Trygve (Journal article; Tidsskriftartikkel; Peer reviewed, 2020-07-14)
      We propose a deep-learning framework to detect and categorize oil spills in synthetic aperture radar (SAR) images at a large scale. Through a carefully designed neural network model for image segmentation trained on an extensive dataset, we obtain state-of-the-art performance in oil spill detection, achieving results that are comparable to results produced by human operators. We also introduce a ...
    • Satellite-Based National Intertidal-Zone Mapping of Continental Norway with Sentinel-1&2 

      Haarpaintner, Jörg; Davids, Corine; Hindberg, Heidi; Arntzen, Ingar M; Borch, Njål Trygve (Research report; Forskningsrapport, 2021-05-21)
      The report describes updated methods that were originally developed in Haarpaintner & Davids (2020) to map the intertidal zone, in terms of atmospheric exposure, type and areal extent, based on radar and optical high resolution (10m) satellite imagery from Sentinel-1A/B (C-band synthetic aperture radar, C-SAR) and Sentinel-2A/B (multi-spectral instruments) of the European Copernicus Program. It ...
    • UX-based personalization of timed media experiences 

      Arntzen, Ingar M; Borch, Njål Trygve (Conference object; Konferansebidrag, 2021)
      A client-side approach brings great opportunities for flexible personalization of timed media experiences. This, turns it into a challenge of UX development. However, UX development does not support time-driven rendering or shared interactivity, which are required by media related scenarios. Moreover, UX development is already quite complicated, so adding support for timed rendering and shared ...