Browsing by Author "Rebriks, Aleksandrs"

Now showing 1 - 2 of 2

Emotion Recognition using EEG signal data from EMO2018 Dataset
(Tartu Ülikool, 2024) Rebriks, Aleksandrs; Avots, Egils, juhendaja; Juuse, Liina, juhendaja; Tartu Ülikool. Loodus- ja täppisteaduste valdkond; Tartu Ülikool. Tehnoloogiainstituut
Emotion Recognition (ER) is developing area within the artificial intelligence field that is focused on comprehending and further interpreting of human emotions through various modalities. Despite that, these approaches are often not ubiquitous as they are affected by external factors. With recent physiology research connecting development of emotions to the central nervous system, usage of brain signals became a highly practical option for emotion recognition. One of the most promising methods of emotion recognition using brain signals for emotion recognition involves using Electroencephalography (EEG). Despite being more complex than classical machine learning or deep learning approaches, EEG-based emotion recognition is potentially more accurate and robust, with applications in mental health monitoring, researches in applied physiology or human-computer interactions. This thesis studies existing approaches of EEG-based emotion recognition methods for private EMO2018 dataset. We adopted methods of Fast Fourier Transform with additional processing for key features extraction and tested different Deep Learning models. Our results show performances of utilized Deep learning models with best accuracy of 88.6% from Hybrid Neural Network approach.
Style transfer based artistic transformation using Albert Gulk’s pencil drawings
(2021) Rebriks, Aleksandrs
Style transfer is the process of retouching images to appear in the artistic style of another image. Researchers have conducted intensive work using advanced machine learning methods and high level mathematical models in order to transfer style of artworks. One of the challenges, which still exists, is to perform style transfer which preserves the shapes of the content image as much as possible while transferring finer style patterns. In this thesis, we investigate how state-ofthe- art style transfer models can be improved such that the shape of the content image is less disturbed during the process. For this purpose we introduce a discrete wavelet transform based neural style transfer pipeline. Experiments were conducted using the artworks of Albert Gulk, a famous Estonian artist of the Kursi school. A challenge in Albert Gulk’s pencil drawings is that they are monotone works, hence using any basic style transfer will significantly disturb the shape of content image. Experimental results show that the proposed wavelet based neural style transfer approach can preserve the shape of content when monotone artworks are used as style images.