Collaborative Research: SaTC: CORE: Medium: Detection of Images and Videos Created by Artificial Intelligence

University Of North Carolina At Charlotte, Charlotte NC

Investigators

Abstract

Artificial intelligence (AI) can generate images and videos that not only can seem real to human senses, but they can also fool other computer programs. AI-generated content are imposing a range of cybersecurity concerns. The objective of this project is to design detectors that are capable of assessing the integrity of digital content. The research will benefit society by providing a deeper understanding about relevant machine learning techniques and ensuring the authenticity of visual and audio content for digital forensics as well as defending users against social engineering attacks. The project team consists of two researchers with expertise in image processing and cybersecurity. The project advances the state of the art in AI-generated visual and audio content detection. The uniqueness of the proposed system is its ability of self-learning and self-evolving to capture content generated by AI models emerging in the future. The proposed mechanisms will allow the detector to adapt to new types of AI-generated images or videos with only a small number of samples, overcoming the limitation of limited samples in existing data-hungry learning algorithms. The proposed defensive mechanisms will ensure the robustness of the detector and prevent it from being attacked by data poisoning or adversarial inputs. The proposed lifelong learning mechanism will enable the detector to leverage accumulated knowledge to achieve self-improvement over time. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.