Now showing items 1-20 of 597

    • A Theoretical Analysis of Deep Neural Networks and Parametric PDEs 

      Kutyniok, Gitta Astrid Hildegard; Petersen, Philipp; Raslan, Mones; Schneider, Reinhold (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-06-02)
      We derive upper bounds on the complexity of ReLU neural networks approximating the solution maps of parametric partial differential equations. In particular, without any knowledge of its concrete shape, we use the inherent low dimensionality of the solution manifold to obtain approximation rates which are significantly superior to those provided by classical neural network approximation results. ...
    • Self-Constructing Graph Convolutional Networks for Semantic Labeling 

      Liu, Qinghui; Kampffmeyer, Michael; Jenssen, Robert; Salberg, Arnt-Børre (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-02-17)
      Graph Neural Networks (GNNs) have received increasing attention in many fields. However, due to the lack of prior graphs, their use for semantic labeling has been limited. Here, we propose a novel architecture called the Self-Constructing Graph (SCG), which makes use of learnable latent variables to generate embeddings and to self-construct the underlying graphs directly from the input features ...
    • Multi-View Self-Constructing Graph Convolutional Networks With Adaptive Class Weighting Loss for Semantic Segmentation 

      Liu, Qinghui; Kampffmeyer, Michael; Jenssen, Robert; Salberg, Arnt Børre (Conference object; Konferansebidrag, 2020-07-28)
      We propose a novel architecture called the Multi-view Self-Constructing Graph Convolutional Networks (MSCG-Net) for semantic segmentation. Building on the recently proposed Self-Constructing Graph (SCG) module, which makes use of learnable latent variables to self-construct the underlying graphs directly from the input features without relying on manually built prior knowledge graphs, we leverage ...
    • Steepening Plasma Density Spectra in the Ionosphere: The Crucial Role Played by a Strong E-Region 

      Ivarsen, Magnus Fagernes; St-Maurice, Jean-Pierre; Jin, Yaqi; Park, Jaeheung; Miloch, Wojciech Jacek; Spicher, Andres; Kwak, Young-Sil; Clausen, Lasse B. N. (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-07-28)
      Based on the Swarm 16 Hz Advanced Plasma Density data set, and using the Swarm A satellite, we apply automatic detection of spectral breaks in seven million sampled plasma density power spectra in the high-latitude F-region ionosphere. This way, we survey the presence of plasma irregularity dissipation due to an enhanced E-region conductance, caused both by solar photoionization and particle ...
    • Blind Super-Resolution Approach for Exploiting Illumination Variety in Optical-Lattice Illumination Microscopy 

      Samanta, Krishnendu; Sarkar, Swagato; Acuña, Sebastian; Joseph, Joby; Ahluwalia, Balpreet Singh; Agarwal, Krishna (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-08-19)
      Optical-lattice illumination patterns help in pushing high spatial frequency components of the sample into the optical transfer function of a collection microscope. However, exploiting these high-frequency components require precise knowledge of illumination if reconstruction approaches similar to structured illumination microscopy are employed. Here, we present an alternate blind reconstruction ...
    • Year-around C- and L-band observation around the MOSAiC ice floe with high spatial and temporal resolution 

      Singha, Suman; Johansson, Malin; Spreen, Gunnar; Howell, Stephen; Shin-ichi, Sobue; Davidson, Malcolm (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-10-12)
      In September 2019, the German research icebreaker Polarstern started the largest multidisciplinary Arctic expedition, the MOSAiC (Multidisciplinary drifting Observatory for the Study of Arctic Climate) drift experiment. Being moored to ice floes at high Arctic for a whole year, thus including the winter season, the main goal of the expedition is to better understand and quantify relevant processes ...
    • Observing electric field and neutral wind with EISCAT 3D 

      Stamm, Johann; Vierinen, Juha; Gustavsson, Björn (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-11-16)
      Measurements of height-dependent electric field (E) and neutral wind (u) are important governing parameters of the Earth's upper atmosphere, which can be used to study, for example, how auroral currents close or how energy flows between the ionized and neutral constituents. The new EISCAT 3D (E3D) incoherent scatter radar will be able to measure a three-dimensional ion velocity vector (v) at each ...
    • CryoSat-2 Significant Wave Height in Polar Oceans Derived Using a Semi-Analytical Model of Synthetic Aperture Radar 2011–2019 

      Heorten, Harold; Tsamados, Michel; Armitage, Thomas; Ridout, Andy; Landy, Jack Christopher (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-10-18)
      This paper documents the retrieval of significant ocean surface wave heights in the Arctic Ocean from CryoSat-2 data. We use a semi-analytical model for an idealised synthetic aperture satellite radar or pulse-limited radar altimeter echo power. We develop a processing methodology that specifically considers both the Synthetic Aperture and Pulse Limited modes of the radar that change close to ...
    • Single-shot fringe pattern phase retrieval using improved period-guided bidimensional empirical mode decomposition and Hilbert transform 

      Gocłowski, Paweł; Cywinska, Maria; Ahmad, Azeem; Ahluwalia, Balpreet Singh; Trusiak, Maciej (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-09-17)
      Fringe pattern analysis is the central aspect of numerous optical measurement methods, e.g., interferometry, fringe projection, digital holography, quantitative phase microscopy. Experimental fringe patterns always contain significant features originating from fluctuating environment, optical system and illumination quality, and the sample itself that severely affect analysis outcome. Before the ...
    • A 10-year record of Arctic summer sea ice freeboard from CryoSat-2 

      Dawson, Geoffrey; Landy, Jack Christopher; Tsamados, Michel; Komarov, Alexander S.; Howell, Stephen; Heorten, Harold; Krumpen, Thomas (Journal article; Tidsskriftartikkel; Peer reviewed, 2022-10-29)
      Satellite observations of pan-Arctic sea ice thickness have so far been constrained to winter months. For radar altimeters, conventional methods cannot differentiate leads from meltwater ponds that accumulate at the ice surface in summer months, which is a critical step in the ice thickness calculation. Here, we use over 350 optical and synthetic aperture radar (SAR) images from the summer months ...
    • Polar Lows - Moist Baroclinic Cyclones in Four Different Vertical Wind Shear Environments 

      Stoll, Patrick; Spengler, Thomas; Terpstra, Annick; Graversen, Rune (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-01-15)
      Polar lows are intense mesoscale cyclones that develop in polar marine air masses. Motivated by the large variety of their proposed intensification mechanisms, cloud structure, and ambient sub-synoptic environment, we use self-organising maps to classify polar lows. The method is applied to 370 polar lows in the north-eastern Atlantic, which were obtained by matching mesoscale cyclones from ...
    • High-throughput spatial sensitive quantitative phase microscopy using low spatial and high temporal coherent illumination 

      Ahmad, Azeem; Dubey, Vishesh; Jayakumar, Nikhil; Habib, Anowarul; Butola, Ankit; Nystad, Mona; Acharya, Ganesh; Basnet, Purusotam; Mehta, Dalip Singh; Ahluwalia, Balpreet Singh (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-08-04)
      High space-bandwidth product with high spatial phase sensitivity is indispensable for a single-shot quantitative phase microscopy (QPM) system. It opens avenue for widespread applications of QPM in the field of biomedical imaging. Temporally low coherence light sources are implemented to achieve high spatial phase sensitivity in QPM at the cost of either reduced temporal resolution or smaller field ...
    • Impacts of snow data and processing methods on the interpretation of long-term changes in Baffin Bay early spring sea ice thickness 

      Glissenaar, Isolde; Landy, Jack Christopher; Petty, Alek; Kurtz, Nathan; Stroeve, Julienne C. (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-10-21)
      In the Arctic, multi-year sea ice is being rapidly replaced by seasonal sea ice. Baffin Bay, situated between Greenland and Canada, is part of the seasonal ice zone. In this study, we present a long-term multi-mission assessment (2003–2020) of spring sea ice thickness in Baffin Bay from satellite altimetry and sea ice charts. Sea ice thickness within Baffin Bay is calculated from Envisat, ...
    • Time series cluster kernels to exploit informative missingness and incomplete label information 

      Mikalsen, Karl Øyvind; Ruiz, Cristina Soguero; Bianchi, Filippo Maria; Revhaug, Arthur; Jenssen, Robert (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-02-20)
      The time series cluster kernel (TCK) provides a powerful tool for analysing multivariate time series subject to missing data. TCK is designed using an ensemble learning approach in which Bayesian mixture models form the base models. Because of the Bayesian approach, TCK can naturally deal with missing values without resorting to imputation and the ensemble strategy ensures robustness to hyperparameters, ...
    • Hyperspectral image classification based on a shuffled group convolutional neural network with transfer learning 

      Liu, Yao; Gao, Lianru; Xiao, Chenchao; Qu, Ying; Zheng, Ke; Marinoni, Andrea (Journal article; Tidsskriftartikkel; Peer reviewed, 2020-06-01)
      Convolutional neural networks (CNNs) have been widely applied in hyperspectral imagery (HSI) classification. However, their classification performance might be limited by the scarcity of labeled data to be used for training and validation. In this paper, we propose a novel lightweight shuffled group convolutional neural network (abbreviated as SG-CNN) to achieve efficient training with a limited ...
    • Evaluation of a sub-kilometre NWP system in an Arctic fjord-valley system in winter 

      Valkonen, Teresa Maaria; Stoll, Patrick; Batrak, Yurii; Køltzow, Morten Andreas Ødegaard; Schneider, Thea Maria; Stigter, Emmy E.; Aashamar, Ola B.; Støylen, Eivind; Jonassen, Marius Opsanger (Journal article; Tidsskriftartikkel; Peer reviewed, 2020-10-28)
      Terrain challenges the prediction of near-surface atmospheric conditions, even in kilometre-scale numerical weather prediction (NWP) models. In this study, the ALADIN-HIRLAM NWP system with 0.5 km horizontal grid spacing and an increased number of vertical levels is compared to the 2.5-km model system similar to the currently operational NWP system at the Norwegian Meteorological Institute. The ...
    • Free-standing tantalum pentoxide waveguides for gas sensing in the mid-infrared 

      Vlk, Marek; Datta, Anurup; Alberti, Sebastian; Murugan, Ganapathy Senthil; Aksnes, Astrid; Jagerska, Jana (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-08-23)
      Typical applications of integrated photonics in the mid-infrared (MIR) are different from near-infrared (telecom) range and, in many instances, they involve chemical sensing through MIR spectroscopy. Such applications necessitate tailored designs of optical waveguides. Both cross-sectional designs and processing methods of MIR waveguides have been a subject of extensive research, where material ...
    • Semi-supervised target classification in multi-frequency echosounder data 

      Choi, Changkyu; Kampffmeyer, Michael; Handegard, Nils Olav; Salberg, Arnt Børre; Brautaset, Olav; Eikvil, Line; Jenssen, Robert (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-08-12)
      Acoustic target classification in multi-frequency echosounder data is a major interest for the marine ecosystem and fishery management since it can potentially estimate the abundance or biomass of the species. A key problem of current methods is the heavy dependence on the manual categorization of data samples. As a solution, we propose a novel semi-supervised deep learning method leveraging a few ...
    • Study of waveguide background at visible wavelengths for on-chip nanoscopy 

      Coucheron, David Andre; Helle, Øystein I.; Wilkinson, James S.; Murugan, Ganapathy Senthil; Domínguez, Carlos; Angelskår, Hallvard; Ahluwalia, Balpreet S. (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-06-17)
      On-chip super-resolution optical microscopy is an emerging field relying on waveguide excitation with visible light. Here, we investigate two commonly used high-refractive index waveguide platforms, tantalum pentoxide (Ta<sub>2</sub>O<sub>5</sub>) and silicon nitride (Si<sub>3</sub>N<sub>4</sub>), with respect to their background with excitation in the range 488–640 nm. The background strength from ...
    • Measurement of snow water equivalent using drone-mounted ultra-wide-band radar 

      Jenssen, Rolf Ole R.; Jacobsen, Svein Ketil (Journal article; Tidsskriftartikkel; Peer reviewed, 2021-07-02)
      The use of unmanned aerial vehicle (UAV)-mounted radar for obtaining snowpack parameters has seen considerable advances over recent years. However, a robust method of snow density estimation still needs further development. The objective of this work is to develop a method to reliably and remotely estimate snow water equivalent (SWE) using UAV-mounted radar and to perform initial field experiments. ...