researcher: Pierre Moulin (director)
A variety of applications in the areas of trust, security, and related areas share a common requirement: advanced data analysis methods are needed to make inferences about the past state of a system. Depending on the application area, the past state of the system may represent the characterization of
- anomalous activity and patterns;
- untrustworthy nodes in a network;
- intrusions in a computer system or in a physical facility;
- chemical or biological contamination; and/or
- tampering with computer hardware, software, data files, audio, video, and other signals, as well as the devices that acquire and display these signals.
Other important applications go beyond the traditional realm of engineering and may make it necessary to make inferences about the health, behavior, or state of mind of a person, and the evolution of a human organization.
Such problems may be broadly described as information forensics problems. The mission of this activity is to provide an intellectual framework that fosters collaboration in that area among researchers from varied backgrounds and thereby increases the potential for breakthroughs.
Participants in this activity conduct research in the abovementioned areas, with a focus on core scientific problems and grand scientific challenges that cannot be solved via the traditional single-discipline research mode. The researchers have a broad range of expertise that includes pattern recognition, machine learning, data mining, detection, and estimation theory; game theory; cryptography; information theory and networking; signal, language, image, and video analysis; computer vision; device modeling; software engineering; chemistry, biology, and life sciences; biometrics; and psychology.
Current research projects in the Information Forensics activity are funded by grants from the federal government and industry partners.