Browsing Institutt for matematiske fag by Author "Lee, Daesoo"

Now showing items 1-5 of 5

    • Ensemble and Self-supervised Learning for Improved Classification of Seismic Signals from the Åknes Rockslope 

      Lee, Daesoo; Aune, Erlend; Langet, Nadege; Eidsvik, Jo (Journal article; Peer reviewed, 2022)
      A case study with seismic geophone data from the unstable Åknes rock slope in Norway is considered. This rock slope is monitored because there is a risk of severe flooding if the massive-size rock falls into the fjord. The ...

    • Musical Source Separation using Diffusion Models 

      Solberg, Aleksander Johnsen (Master thesis, 2023)
      Musikalsk kildeseparering er en utfordrende oppgave innen lydsignalbehandling som har som mål å ekstrahere individuelle kilder fra en musikalsk blanding, for eksempel å separere vokal fra bakgrunnsmusikk. I de siste årene ...

    • Standardization in building an ANN-based mooring line top tension prediction system 

      Lee, Daesoo; Lee, Seungjae; Lee, Jaeyong (Peer reviewed; Journal article, 2022)
      ANN-based mooring line top tension prediction systems are trained on ship motion and mooring line top tension time histories from multiple wave states with a certain simulation length. In the previous studies, selection ...

    • Vector Quantized Time Series Generation using Diffusion Models 

      Tufte, Martin Gudahl (Master thesis, 2023)
      Tidsrekkegenerasjon (TSG) er en type generativ modellering som fokuserer på å lære fordelingen av tidsrekker. En ny tilnærming til TSG, kalt TimeVQVDM, er foreslått for å kombinere vektorkvantisering med diffusjonsmodell ...

    • Vector Quantized Time Series Generation with a Bidirectional Prior Model 

      Lee, Daesoo; Malacarne, Sara; Aune, Erlend (Journal article; Peer reviewed, 2023)
      Time series generation (TSG) studies have mainly focused on the use of Generative Adversarial Networks (GANs) combined with recurrent neural network (RNN) variants. However, the fundamental limitations and challenges of ...