Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
2 results
Search Results
Item Restricted Deepfake Face Images Detection(Bahrain Polytechnic, 2024) Aldalbahi, Bedour Ahmad; Fawzy, Abdelhameed IbrahimDeepfake is a sort of AI that forges original image or video and create persuading images, audio and video. Deepfake media continues to gain ground online, raising a number of ethical and moral questions about their use, in that deepfakes can be used to undermine political elections, companies, individual and corporate finances, reputation, and many more. The proposed system to solve this problem is to use the most popular algorithm in deep learning, Convolution Neural Network (CNN), for detecting fake images. This will be achieved by training two deep learning models and analyzing their performances in distinguishing between the two classes of images “Real”,” Fake”. Our main aim is to contribute a useful framework toward the detection of deep-fake photos with deep learning. This thesis proposed convolutional neural networks for the identification of genuine and deepfake pictures. In this study, we have trained two models: DenseNet121 and ResNet50. The results will be categorized by Four evaluation metrics: accuracy, precision, recall, and F1-score. In that respect, DenseNet121 had the best performance with an accuracy of 94%. Besides, we obtained 91% from the ResNet50.8 0Item Restricted Security of Text to Image Conversions(2023-07-27) Alssadi, Zobaida; Silaghi, MariusThe use of images and icons to represent news or narratives has grown in popularity. Still, one critical problem is that they are not equivalent to language, making them vulnerable to adversary attacks. This study examines the impact of image-poisoning attacks based on polysemantic words and of image attacks based on cultural differences when converting text to images. Such attacks can lead to the loss of important information and create confusion and incorrect interpretations of the intended meaning, misinforming the general public. The study specifically focuses on possible effects in a news and story context. This study highlights the significance of taking security considerations into account when image-based attacks are relevant and motivates the development of strategies to ensure that information is conveyed through images and icons in a culturally appropriate and accurate manner, as well as to prevent image tampering and the spread of false information by attackers.21 0