|
|
Self-Supervised Learning for Recognition of Sports Poses in Image
Olekšák, Samuel ; Kocur, Viktor (referee) ; Herout, Adam (advisor)
This thesis demonstrates a solution for minimizing the amount of necessary labelled training data in the classification of sports poses using a neural network trained with contrastive self-supervised learning. Training consists of two stages. The first stage trains a feature extractor which uses unlabelled training images extracted from recordings of exercises from multiple viewpoints. In the second stage, using a small amount of labelled data, a simple classifier connected to the feature extractor is trained. The thesis discusses classification in the context of yoga poses, however, the final solution can be easily applied to any other sport in case of obtaining a suitable dataset. During the development of the solution, emphasis is placed on the performance of the resulting model so that it can be used on mobile devices. The resulting model reached an accuracy of 76 % using augmentations with a data set containing four labelled images per yoga pose. On a larger data set with 800 labelled images for all poses, an accuracy of 82 % is reached.
|
|
|
Emotion Recognition from Brain Electroencephalogram (EEG) Signals
Fritz, Karel ; Jawed, Soyiba (referee) ; Malik, Aamir Saeed (advisor)
Tato studie se zaměřuje na klasifikaci emocí z elektroencefalogramu (EEG). Kombinuje znalosti o fyziologii mozku (a emocí), s frekvenční analýzou, analýzou složitosti, zpracov- áním signálů a hlubokým strojovým učením (CNN, GNN). Cílem této práce je vytvořit model pro klasifikaci emocí a poskytnout nové náhledy do rozpoznávání emocí z EEG. Vytvořené modely stojí na principech CNN, GNN, multitask a self supervised tréninku. Jedním z výsledků bylo dosažení State of the Art výsledků na datasetu SEED. Proces porozumění této úloze sdílím na konci této práce.
|
|
|
Identifikace osob pomocí obrazu duhovky
Žákovic, Marek ; Hradiš, Michal (referee) ; Vaško, Marek (advisor)
The goal of this bachelor’s thesis was to create a system for person identification using iris images. The thesis describes existing methods and procedures for iris recognition. The proposed method utilizes a convolutional neural network trained to extract features, which are then used to compare whether the image belongs to the same person or not. The experiments involve training and evaluating the neural network. For the purposes of this thesis, freely available datasets were used, which were modified for specific use.
|
|
|
High Level Analysis of the Psychotherapy Sessions
Polok, Alexander ; Karafiát, Martin (referee) ; Matějka, Pavel (advisor)
This work focuses on analyzing psychotherapy sessions within the DeePsy research project. This work aims to design and develop features that model the session dynamics, which can reveal seemingly subtle nuances. The mentioned features are automatically extracted from the source recording using neural networks. They are further processed, compared across sessions, and displayed graphically, creating a document that acts as a feedback document about the session for the therapist. Furthermore, this assistive tool can help therapists to professionally grow and to provide better psychotherapy in the future. A relative improvement in voice activity detection of 37.82% was achieved. The VBx diarization system was generalized to converge to two speakers with a minimum relative error rate degradation of 0.66%. An automatic speech recognition system has been trained with a 17.06% relative improvement over the best available hybrid model. Models for sentiment classification, type of therapeutic interventions, and overlapping speech detection were also trained.
|
|
|
Self-Supervised Learning for Recognition of Sports Poses in Image
Olekšák, Samuel ; Kocur, Viktor (referee) ; Herout, Adam (advisor)
This thesis demonstrates a solution for minimizing the amount of necessary labelled training data in the classification of sports poses using a neural network trained with contrastive self-supervised learning. Training consists of two stages. The first stage trains a feature extractor which uses unlabelled training images extracted from recordings of exercises from multiple viewpoints. In the second stage, using a small amount of labelled data, a simple classifier connected to the feature extractor is trained. The thesis discusses classification in the context of yoga poses, however, the final solution can be easily applied to any other sport in case of obtaining a suitable dataset. During the development of the solution, emphasis is placed on the performance of the resulting model so that it can be used on mobile devices. The resulting model reached an accuracy of 76 % using augmentations with a data set containing four labelled images per yoga pose. On a larger data set with 800 labelled images for all poses, an accuracy of 82 % is reached.
|
guest ::
