"autoencoder" - Selaus asiasanan mukaan Kandidaatintutkielmat

Viitteet 1-2 / 2

    • The Effect of Feature Space Topology on Speech Emotion Recognition 

      Niskanen, Joel (2024)
      Kandidaatintyö
      The main purpose of speech emotion recognition (SER) is to determine the speaker’s current emotional state automatically. Typically, various machine learning models have been used to approach this problem. However, the ...

    • Visualization of Semantic Segmentation Networks 

      Kankainen, Ossi (2020)
      Kandidaatintyö
      The development of visualization methods for deep convolutional neural networks supports their design and helps in their adaption also to critical applications. Semantic segmentation has many such heavily regulated application ...