Automated 3D facial landmarks localization for 4D dataset
dc.contributor.author | Papaioannou, Panagiotis | en |
dc.date.accessioned | 2020-06-09T13:09:35Z | |
dc.date.available | 2020-06-10T00:00:40Z | |
dc.date.issued | 2020-06-09 | |
dc.identifier.uri | https://repository.ihu.edu.gr//xmlui/handle/11544/29528 | |
dc.rights | Default License | |
dc.subject | 3D | en |
dc.subject | Facial landmarks | en |
dc.title | Automated 3D facial landmarks localization for 4D dataset | en |
heal.abstract | 3D facial landmarks are crucial as a first step for various applications like alignment and registration which implies that successful localization can lead to better outcomes for further approaches. Even though the problem of landmarks localization has been examined in depth for the case of 2D images, the amount of work that has been done for the 3D case is significantly smaller and it can be said that is negligible for the 4D case (3D plus time). In this thesis, a literature review for the current approaches for 3D a localization of facial landmarks is carried out. Finally, experiments of a new method show validity and improvement in certain landmarks. | en |
heal.academicPublisher | IHU | en |
heal.academicPublisherID | ihu | en_US |
heal.access | free | en_US |
heal.advisorName | Kotsakis, Rigas | en |
heal.classification | Machine Learning | en |
heal.committeeMemberName | Kotsakis, Rigas | en |
heal.committeeMemberName | Baltatzis, Dimitrios | en |
heal.committeeMemberName | Chatzimisios, Periklis | en |
heal.creatorID.dhareID | panos_wane@hotmail.com | |
heal.language | en | en_US |
heal.license | http://creativecommons.org/licenses/by-nc/4.0 | en_US |
heal.publicationDate | 2020-05-22 | |
heal.recordProvider | School of Science and Technology, MSc in Data Science | en_US |
heal.type | masterThesis | en_US |
Αρχεία
Πρωτότυπος φάκελος/πακέτο
1 - 1 από 1
Φόρτωση...
- Ονομα:
- Dissertation!final!cor..pdf
- Μέγεθος:
- 1.15 MB
- Μορφότυπο:
- Adobe Portable Document Format
- Περιγραφή:
Φάκελος/Πακέτο αδειών
1 - 1 από 1
Δεν υπάρχει διαθέσιμη μικρογραφία
- Ονομα:
- license.txt
- Μέγεθος:
- 2.58 KB
- Μορφότυπο:
- Item-specific license agreed upon to submission
- Περιγραφή: