dc.contributor.author |
Mohammed Shiyam, Abdul baasith |
|
dc.date.accessioned |
2024-03-04T03:50:40Z |
|
dc.date.available |
2024-03-04T03:50:40Z |
|
dc.date.issued |
2023 |
|
dc.identifier.citation |
Mohammed Shiyam, Abdul baasith (2023) Deepfake Low Resource Image Detection with Explainable Reporting. BSc. Dissertation, Informatics Institute of Technology |
en_US |
dc.identifier.issn |
2019566 |
|
dc.identifier.uri |
http://dlib.iit.ac.lk/xmlui/handle/123456789/1807 |
|
dc.description.abstract |
"In this research, the author proposes a solution for deepfake detection in profile images using deep
learning approaches, including transfer learning and hybrid models, along with Explainable AI
(XAI). The proposed solution leverages transfer learning to fine-tune a pre-trained deep neural
network and alter the architecture of models to improve the overall performance. Additionally,
XAI is employed to increase the interpretability and transparency of the decision-making process,
enabling the understanding of why a particular image was classified as fake or real. The proposed solution was evaluated on a on various testing matrix and was able achieved high accuracy,
robustness, and interpretability. These results demonstrate the potential of transfer learning, hybrid models, and XAI for effective deepfake detection in profile images." |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
IIT |
en_US |
dc.subject |
Deep Learning |
en_US |
dc.subject |
Image Processing |
en_US |
dc.subject |
Explainable AI |
en_US |
dc.title |
Deepfake Low Resource Image Detection with Explainable Reporting |
en_US |
dc.type |
Thesis |
en_US |