Neural network with agnostic meta-learning model for face-aging recognition / Rasha Ragheb Attaallah

Rasha Ragheb , Attaallah (2022) Neural network with agnostic meta-learning model for face-aging recognition / Rasha Ragheb Attaallah. PhD thesis, Universiti Malaya.

[img] PDF (Thesis PhD)
Download (1762Kb)

    Abstract

    Face recognition is one of the most popular and accessible verification techniques. It is also accepted by users as it is non-invasive. Nevertheless, the aging process may change the face shape and texture. Therefore, aging is considered one factor that affects the accuracy of face recognition applications. Existing techniques and methods for face aging recognition are degraded in performance due to many factors, such as the uncontrolled nature of aging processes of the human face. Thus, there is a need to further investigate face aging recognition techniques, particularly the one which can be used to compare two images of faces for the same person at a different age. This study aims to design and model a framework to recognize face aging based on artificial neural networks and Model-agnostic meta-learning (MAML), using parameters obtained from identical tasks with certain updates on these parameters. The thesis presents the main parts of the research methodology. The literature review and problem extraction are reviewed in phase 1. Phase 2 explains the research objectives. Phase 3 shows the proposed model design and implementation. Finally, phase 4 shows the analysis of the results based on the proposed framework. The methodology starts with phase one, which includes the literature review and problem extraction. During this phase, a review of existing research was done, followed by the extraction of all face recognition challenges. This helps in defining the problem statement. The second phase includes the determination of the research objectives and explaining them. The third phase proposes the framework, shows the design and the implementation. The fourth phase explains the results, demonstrates the analyses and the evalutes the data.

    Item Type: Thesis (PhD)
    Additional Information: Thesis (PhD) – Faculty of Computer Science & Information Technology, Universiti Malaya, 2022.
    Uncontrolled Keywords: Neural network; Agnostic meta-learning model; Face-aging recognition; CMU Multi-PIE; ORL dataset
    Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
    Q Science > QA Mathematics > QA76 Computer software
    Divisions: Faculty of Computer Science & Information Technology > Dept of Information System
    Depositing User: Mr Mohd Safri Tahir
    Date Deposited: 26 May 2025 02:14
    Last Modified: 26 May 2025 02:14
    URI: http://studentsrepo.um.edu.my/id/eprint/15243

    Actions (For repository staff only : Login required)

    View Item