Beridze, Jevgenija;
(2021)
A PCA approach to the object constancy for faces using view-based models of the face.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
Beridze_PhD_thesis.pdf Download (7MB) | Preview |
Abstract
The analysis of object and face recognition by humans attracts a great deal of interest, mainly because of its many applications in various fields, including psychology, security, computer technology, medicine and computer graphics. The aim of this work is to investigate whether a PCA-based mapping approach can offer a new perspective on models of object constancy for faces in human vision. An existing system for facial motion capture and animation developed for performance-driven animation of avatars is adapted, improved and repurposed to study face representation in the context of viewpoint and lighting invariance. The main goal of the thesis is to develop and evaluate a new approach to viewpoint invariance that is view-based and allows mapping of facial variation between different views to construct a multi-view representation of the face. The thesis describes a computer implementation of a model that uses PCA to generate example- based models of the face. The work explores the joint encoding of expression and viewpoint using PCA and the mapping between viewspecific PCA spaces. The simultaneous, synchronised video recording of 6 views of the face was used to construct multi-view representations, which helped to investigate how well multiple views could be recovered from a single view via the content addressable memory property of PCA. A similar approach was taken to lighting invariance. Finally, the possibility of constructing a multi-view representation from asynchronous view-based data was explored. The results of this thesis have implications for a continuing research problem in computer vision – the problem of recognising faces and objects from different perspectives and in different lighting. It also provides a new approach to understanding viewpoint invariance and lighting invariance in human observers.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | A PCA approach to the object constancy for faces using view-based models of the face |
| Event: | UCL (University College London) |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| Additional information: | Copyright © The Author 2021. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Brain Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Brain Sciences > Div of Psychology and Lang Sciences |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10132817 |
Archive Staff Only
 |
View Item |
