Institutional Repository of Vilnius University
Emocijų atpažinimas balse giluminių neuroninių tinklų pagalba /
Abstract
dc:descriptionThis work studies the use of deep neural networks for recognizing voice emotions: fear, anger, joy, surprise, disgust, sadness and neutral. ConvNeXtTiny, RasNet50, EfficientNetV2S convolutional neural network models were tested, with EfficientNetV2S achieving the best results. Additionally, different data augmentation methods were evaluated. The combination of EfficientNetV2S and random time and frequency data masking augmentation method for spectrograms achieved the best results with a final accuracy of 86.69% when classifying emotions.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Truikys, Paulius,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210648133&prefLang=en_US