Supervisor(s): Guangdong Bai

This study focuses on evaluating and identifying potential security vulnerabilities within the ChatGPT plugin system. By analyzing the interaction between plugins,

Supervisor(s): Dan Kim

Advancements in AI and deep learning have improved image and video generation but introduced security risks. This study strengthens deepfake detection by applying multi-scale training and noise augmentation, leading to a 9% performance boost. The results highlight the effectiveness of the proposed approach in enhancing detection across varying resolutions and augmented images.