Biography
Dr. Shu Hu is an assistant professor in the School of Applied and Creative Computing and the Director of the Purdue Machine Learning and Media Forensics (M2) Lab at Purdue University. He was a Post-Doctoral Fellow at the Heinz College of Carnegie Mellon University from 2022 to 2023. He received my Ph.D. degree in Computer Science and Engineering from the University at Buffalo, SUNY in 2022. He is the recipient of the National AI Research Resource (NAIRR) Pilot award (2024), the National Science Foundation CRII Award (2024), the Machine Intelligence Research Outstanding Reviewer Award (2023), and SUNY Buffalo's CSE Best PhD Dissertation Award (2022). His research interests include machine learning, media forensics, and computer vision.
SRP Project Title
Improving Fairness in Detecting AI-Synthesized Fake Multimedia
NAIRR Project
Improving Fairness in Detecting AI-Synthesized Fake Multimedia
Topical Areas
Applied Computer Science; Artificial Intelligence and Intelligent Systems; Computer Science; Media and communications; Other Computer and Information Sciences; Other Engineering and Technologies; Performance Evaluation and Benchmarking; Visualization and Human-Computer Systems
Abstract
DeepFake, a term increasingly mentioned in the news and social media, refers to highly realistic fake images, and videos created using AI algorithms. Combating DeepFake technology requires a comprehensive strategy that extends well beyond the realm of mere detection, emphasizing the responsible design, development, and deployment of technologies. This also known as responsible forensics, focuses on applying forensic science to digital content ethically, ensuring that actions to identify and mitigate DeepFakes meet high ethical standards and respect for human rights. At the heart of responsible forensics lies the commitment to fairness, especially important in the context of generative AI. It is crucial for detection tools to be crafted and used in ways that prevent unintentional bias against certain individuals or groups, thus upholding justice and equality in the digital realm. Therefore, our goal is to improve fairness in detecting novel DeepFakes.
Desired Skills
Python Programming Skill.
Lightning Talk Title
Improving Fairness in Detecting AI-Synthesized Fake Multimedia
Keywords
DeepFake Detection; Media Forensics; Generalization; AI-generated Media