Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Data augmentation methods for Vision Transformers

Georgakilas Christos

Simple record


URIhttp://purl.tuc.gr/dl/dias/E7D90BFF-93A5-4DC0-A9D9-92B95EB1FE51-
Identifierhttps://doi.org/10.26233/heallink.tuc.93681-
Languageen-
Extent2.4 megabytesen
Extent40 pagesen
TitleData augmentation methods for Vision Transformersen
TitleΜέθοδοι επαύξησης δεδομένων για νευρωνικά δίκτυα Vision Transformerel
CreatorGeorgakilas Christosen
CreatorΓεωργακιλας Χριστοςel
Contributor [Thesis Supervisor]Zervakis Michailen
Contributor [Thesis Supervisor]Ζερβακης Μιχαηλel
Contributor [Committee Member]Lagoudakis Michailen
Contributor [Committee Member]Λαγουδακης Μιχαηλel
Contributor [Committee Member]Κομοντάκης Νίκοςel
Contributor [Committee Member]Komodakis Nikosen
PublisherΠολυτεχνείο Κρήτηςel
PublisherTechnical University of Creteen
Academic UnitTechnical University of Crete::School of Electrical and Computer Engineeringen
Academic UnitΠολυτεχνείο Κρήτης::Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστώνel
Content SummaryThe Transformer architecture was first introduced in 2017 and has since become the standard for Natural Language Processing tasks, replacing Recurrent Neural Networks. For the first time, in 2021, the Transformer architecture was used with great success for computer vision tasks, proving that a Vision Transformer can, under certain conditions, outperform Convolutional Neural Networks and become the state-of-the-art in image recognition. One of the main challenges being tackled by subsequent work on Vision Transformers is the need of the architecture for humongous amounts of data during pre-training in order to achieve state-of-the-art accuracy on the downstream task. Some works have addressed this by altering or adding parts to the original Vision Transformer architecture while others are using Self-Supervised Learning techniques to take advantage of unlabeled data. This thesis explores data augmentation methods for Vision Transformers with the goal to increase the model’s accuracy and robustness on classification tasks, with limited amounts of data. Our augmentation methods are based on the architecture’s characteristics such as the self-attention mechanism and the input of discrete tokens. All methods are tested for the benchmark classification datasets CIFAR-10 and CIFAR-100 using Supervised Learning and yield great results. When training with the same model hyperparameters, our best augmentation method improves the baseline’s accuracy on CIFAR-10 and CIFAR-100 by 1.98 % and 2.71 % respectively. en
Type of ItemΔιπλωματική Εργασίαel
Type of ItemDiploma Worken
Licensehttp://creativecommons.org/licenses/by/4.0/en
Date of Item2022-10-17-
Date of Publication2022-
SubjectVision Transformersen
Bibliographic CitationChristos Georgakilas, "Data augmentation methods for Vision Transformers", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2022en
Bibliographic CitationΧρίστος Γεωργακίλας, "Μέθοδοι επαύξησης δεδομένων για νευρωνικά δίκτυα Vision Transformer", Διπλωματική Εργασία, Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών, Πολυτεχνείο Κρήτης, Χανιά, Ελλάς, 2022el

Available Files

Services

Statistics